notebooks currently used

notebooks/.ipynb_checkpoints/CVAE-Copy1-checkpoint.ipynb

@@ -0,0 +1,1833 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "10a00e46-827a-4278-a715-99526591a0a7",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import torch.nn as nn\n",
+    "import torch\n",
+    "class CondVAE(nn.Module):\n",
+    "    def __init__(self, dim_input, latent_dim=10, context_vector_size=6):\n",
+    "        super(CondVAE, self).__init__()\n",
+    "        \n",
+    "        self.latent_dim = latent_dim\n",
+    "\n",
+    "        # Encoder\n",
+    "        self.encoder = nn.Sequential(\n",
+    "            nn.Linear(in_features=dim_input, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=300),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=300, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Flatten()\n",
+    "        )\n",
+    "        \n",
+    "        self.fc_mu = nn.Linear(100, latent_dim)\n",
+    "        self.fc_logvar = nn.Linear(100, latent_dim)\n",
+    "\n",
+    "        # Decoder\n",
+    "        self.decoder = nn.Sequential(\n",
+    "            nn.Linear(in_features=latent_dim+6, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=300),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=300, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=1),\n",
+    "        )\n",
+    "\n",
+    "    def encode(self, x):\n",
+    "        x = self.encoder(x)\n",
+    "        mu = self.fc_mu(x)\n",
+    "        log_var = self.fc_logvar(x)\n",
+    "\n",
+    "        return mu, log_var\n",
+    "    \n",
+    "    def decode(self, z, context_vector):\n",
+    "        # Concatenate the sampling (latent distribution) + embedding -> samples conditioned on both the input data and the specified label\n",
+    "        #print(properties.shape, z.shape)\n",
+    "        zcomb = torch.concat((z, context_vector), 1)\n",
+    "        #print(zcomb.shape)\n",
+    "        \n",
+    "        return self.decoder(zcomb)     \n",
+    "    \n",
+    "    def sampling(self, mu, log_var):\n",
+    "        # calculate standard deviation\n",
+    "        std = log_var.mul(0.5).exp_()\n",
+    "        \n",
+    "        # create noise tensor of same size as std to add to the latent vector\n",
+    "        eps = torch.cuda.FloatTensor(std.size()).normal_()\n",
+    "        \n",
+    "        # multiply eps with std to scale the random noise according to the learned distribution + add combined\n",
+    "        return eps.mul(std).add_(mu) # return z sample \n",
+    "\n",
+    "    def forward(self, x, context_vector):\n",
+    "        mu, log_var = self.encode(x)\n",
+    "        z = self.sampling(mu, log_var)\n",
+    "        #print(z.shape)\n",
+    "\n",
+    "        return self.decode(z, context_vector), mu, log_var\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "8bbbe719-f9a3-4ace-ac4c-ed29ec4f9486",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import tqdm\n",
+    "import torch\n",
+    "import torch.nn as nn\n",
+    "\n",
+    "def condvae_loss(pred, label, mu, logvar):\n",
+    "    \"\"\"\n",
+    "    Calculate the conditional Variational Autoencoder (cVAE) loss.\n",
+    "\n",
+    "    This function computes the cVAE loss, which consists of two components:\n",
+    "    - Reconstruction loss: Measures the discrepancy between the reconstructed\n",
+    "      data and the original input.\n",
+    "    - KL divergence loss: Quantifies the difference between the learned latent\n",
+    "      distribution and the desired prior distribution (Gaussian).\n",
+    "\n",
+    "    Args:\n",
+    "        recon_x (torch.Tensor): Reconstructed data from the VAE.\n",
+    "        x (torch.Tensor): Original input data.\n",
+    "        mu (torch.Tensor): Latent variable mean.\n",
+    "        logvar (torch.Tensor): Logarithm of latent variable variance.\n",
+    "\n",
+    "    Returns:\n",
+    "        torch.Tensor: Computed cVAE loss.\n",
+    "    \"\"\"\n",
+    "    \n",
+    "    # MSE loss element-wise and sums up the individual losses\n",
+    "    cde_loss = nn.MSELoss(reduction='mean')(pred, label)\n",
+    "    \n",
+    "    # quantifies the difference between the learned latent distribution and the desired prior distribution (Gaussian)\n",
+    "    kl_divergence = -0.5 * torch.sum(1 + logvar - mu.pow(2) - logvar.exp())\n",
+    "    \n",
+    "    return cde_loss + kl_divergence\n",
+    "\n",
+    "def VAE_trainEpoch(model, optimizer, train_loader, dim_in=100):\n",
+    "    \"\"\"\n",
+    "    Train a Variational Autoencoder (VAE) for one epoch.\n",
+    "\n",
+    "    This function trains a VAE for one epoch using the provided data loader.\n",
+    "    It calculates the cVAE loss, performs backpropagation, and updates the model's parameters.\n",
+    "\n",
+    "    Args:\n",
+    "        model (nn.Module): VAE model to be trained.\n",
+    "        optimizer (torch.optim.Optimizer): Optimizer for updating model parameters.\n",
+    "        train_loader (DataLoader): DataLoader containing training data.\n",
+    "        dim_in (int): Dimensionality of the input noise.\n",
+    "\n",
+    "    Returns:\n",
+    "        float: Average loss for the epoch.\n",
+    "    \"\"\"\n",
+    "    model = model.train()\n",
+    "    device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
+    "    total_loss = 0\n",
+    "\n",
+    "    progress_bar = tqdm.tqdm(train_loader, desc=\"Epoch Progress\", leave=False)\n",
+    "    for context_vector, label in progress_bar:\n",
+    "        context_vector = context_vector.to(device)\n",
+    "        datain = torch.randn(size=(len(context_vector), dim_in)).to(device)\n",
+    "        label = label.unsqueeze(1).cuda()\n",
+    "        optimizer.zero_grad()\n",
+    "\n",
+    "        recon_batch, mu, log_var = model(datain, context_vector)\n",
+    "        loss = condvae_loss(recon_batch, label, mu, log_var)\n",
+    "\n",
+    "        loss.backward()\n",
+    "        optimizer.step()\n",
+    "\n",
+    "        total_loss += loss.item()\n",
+    "        progress_bar.set_postfix({\"Loss\": total_loss / (progress_bar.n + 1)})\n",
+    "\n",
+    "    return total_loss / len(train_loader)\n",
+    "\n",
+    "def VAE_train(model, optimizer, train_loader, epochs, dim_in, save_path=None):\n",
+    "    \"\"\"\n",
+    "    Train a Variational Autoencoder (VAE) for multiple epochs.\n",
+    "\n",
+    "    This function trains a VAE for the specified number of epochs using the provided data loader.\n",
+    "    It prints the epoch progress and the computed loss for each epoch.\n",
+    "\n",
+    "    Args:\n",
+    "        model (nn.Module): VAE model to be trained.\n",
+    "        optimizer (torch.optim.Optimizer): Optimizer for updating model parameters.\n",
+    "        train_loader (DataLoader): DataLoader containing training data.\n",
+    "        epochs (int): Number of epochs for training.\n",
+    "\n",
+    "    Returns:\n",
+    "        None\n",
+    "    \"\"\"\n",
+    "    for epoch in range(epochs):\n",
+    "        print(f\"Epoch {epoch + 1}/{epochs}\")\n",
+    "        epoch_loss = VAE_trainEpoch(model, optimizer, train_loader, dim_in)\n",
+    "        print(f\"Epoch {epoch + 1}/{epochs}, Loss: {epoch_loss:.4f}\")\n",
+    "        \n",
+    "    if save_path!=None:\n",
+    "        torch.save(model, save_path)\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "11aaa0e9-e745-483a-887d-d851e791f8e4",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "from astropy.io import fits\n",
+    "import os\n",
+    "from astropy.table import Table\n",
+    "from scipy.spatial import KDTree\n",
+    "\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "from IPython.display import Image\n",
+    "from IPython.core.display import HTML "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "0814440a-e341-4540-bfba-466b74b9873d",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import torch\n",
+    "from torch.utils.data import DataLoader, dataset, TensorDataset\n",
+    "from torch import nn, optim\n",
+    "from torch.optim import lr_scheduler"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "8ecf9b60-bd03-4fa6-9516-c767d04b2071",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import sys\n",
+    "sys.path.append('../insight')\n",
+    "from archive import archive \n",
+    "from insight_arch import Photoz_network\n",
+    "from insight import Insight_module\n",
+    "from utils import sigma68, nmad, plot_photoz_estimates\n",
+    "from scipy import stats"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 272,
+   "id": "1277097d-e4bb-4bdd-b1f4-bccc88be0169",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "from matplotlib import rcParams\n",
+    "rcParams[\"mathtext.fontset\"] = \"stix\"\n",
+    "rcParams[\"font.family\"] = \"STIXGeneral\"\n",
+    "parent_dir = '/data/astro/scratch/lcabayol/Euclid/NNphotozs/Euclid_EXT_MER_PHZ_DC2_v1.5'"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 273,
+   "id": "5d2d3713-ff7f-4f16-860f-cf5ff42801b1",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "photoz_archive = archive(path = parent_dir, Qz_cut=1)\n",
+    "f, ferr, specz, specqz = photoz_archive.get_training_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 274,
+   "id": "45af2d9e-1160-4859-9888-f5daf62df84a",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "dset = TensorDataset(torch.Tensor(f),torch.Tensor(specz))\n",
+    "loader = DataLoader(dset, batch_size=100, shuffle=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 275,
+   "id": "dad8733c-c36a-4e86-b32d-41c244ba6259",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "dim_input=50\n",
+    "latent_dim=10\n",
+    "context_vector_dim=6\n",
+    "epochs=100\n",
+    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 276,
+   "id": "ef0b6560-dd3d-4b9d-a14c-43bf9338f7a4",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "vae = CondVAE(dim_input, latent_dim=latent_dim, context_vector_size=6).to(device)\n",
+    "optimizer = optim.Adam(vae.parameters(), lr=1e-3, weight_decay=1e-4)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "dd04d484-d14d-488d-a640-180fb6ab1001",
+   "metadata": {
+    "collapsed": true,
+    "jupyter": {
+     "outputs_hidden": true
+    },
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 1/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                            \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 1/100, Loss: 70.1670\n",
+      "Epoch 2/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                            \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 2/100, Loss: 5.8741\n",
+      "Epoch 3/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 3/100, Loss: 0.4596\n",
+      "Epoch 4/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 4/100, Loss: 0.9228\n",
+      "Epoch 5/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 5/100, Loss: 0.5939\n",
+      "Epoch 6/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 6/100, Loss: 0.7836\n",
+      "Epoch 7/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 7/100, Loss: 0.4829\n",
+      "Epoch 8/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 8/100, Loss: 0.5570\n",
+      "Epoch 9/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 9/100, Loss: 0.4176\n",
+      "Epoch 10/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                            \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 10/100, Loss: 0.9913\n",
+      "Epoch 11/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 11/100, Loss: 0.3367\n",
+      "Epoch 12/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 12/100, Loss: 0.3655\n",
+      "Epoch 13/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 13/100, Loss: 0.2642\n",
+      "Epoch 14/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 14/100, Loss: 0.2729\n",
+      "Epoch 15/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 15/100, Loss: 0.2490\n",
+      "Epoch 16/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 16/100, Loss: 0.2426\n",
+      "Epoch 17/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 17/100, Loss: 0.2418\n",
+      "Epoch 18/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 18/100, Loss: 0.2341\n",
+      "Epoch 19/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 19/100, Loss: 0.2288\n",
+      "Epoch 20/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 20/100, Loss: 0.2216\n",
+      "Epoch 21/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 21/100, Loss: 0.2200\n",
+      "Epoch 22/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 22/100, Loss: 0.2140\n",
+      "Epoch 23/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 23/100, Loss: 0.2125\n",
+      "Epoch 24/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 24/100, Loss: 0.2178\n",
+      "Epoch 25/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 25/100, Loss: 0.2127\n",
+      "Epoch 26/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 26/100, Loss: 0.2057\n",
+      "Epoch 27/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 27/100, Loss: 0.2168\n",
+      "Epoch 28/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 28/100, Loss: 0.2043\n",
+      "Epoch 29/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 29/100, Loss: 0.2039\n",
+      "Epoch 30/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 30/100, Loss: 0.1975\n",
+      "Epoch 31/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 31/100, Loss: 0.1956\n",
+      "Epoch 32/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 32/100, Loss: 0.1958\n",
+      "Epoch 33/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 33/100, Loss: 0.1893\n",
+      "Epoch 34/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 34/100, Loss: 0.1890\n",
+      "Epoch 35/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 35/100, Loss: 0.1869\n",
+      "Epoch 36/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 36/100, Loss: 0.1818\n",
+      "Epoch 37/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 37/100, Loss: 0.1784\n",
+      "Epoch 38/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 38/100, Loss: 0.1761\n",
+      "Epoch 39/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 39/100, Loss: 0.1757\n",
+      "Epoch 40/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 40/100, Loss: 0.1746\n",
+      "Epoch 41/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 41/100, Loss: 0.1756\n",
+      "Epoch 42/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 42/100, Loss: 0.1711\n",
+      "Epoch 43/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 43/100, Loss: 0.1706\n",
+      "Epoch 44/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 44/100, Loss: 0.1681\n",
+      "Epoch 45/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 45/100, Loss: 0.1672\n",
+      "Epoch 46/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 46/100, Loss: 0.1617\n",
+      "Epoch 47/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 47/100, Loss: 0.1637\n",
+      "Epoch 48/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 48/100, Loss: 0.1681\n",
+      "Epoch 49/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 49/100, Loss: 0.1637\n",
+      "Epoch 50/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 50/100, Loss: 0.1584\n",
+      "Epoch 51/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 51/100, Loss: 0.1591\n",
+      "Epoch 52/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 52/100, Loss: 0.1583\n",
+      "Epoch 53/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 53/100, Loss: 0.1536\n",
+      "Epoch 54/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 54/100, Loss: 0.1584\n",
+      "Epoch 55/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 55/100, Loss: 0.1624\n",
+      "Epoch 56/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 56/100, Loss: 0.1602\n",
+      "Epoch 57/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 57/100, Loss: 0.1547\n",
+      "Epoch 58/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 58/100, Loss: 0.1540\n",
+      "Epoch 59/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 59/100, Loss: 0.1541\n",
+      "Epoch 60/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 60/100, Loss: 0.1505\n",
+      "Epoch 61/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 61/100, Loss: 0.1521\n",
+      "Epoch 62/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 62/100, Loss: 0.1504\n",
+      "Epoch 63/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 63/100, Loss: 0.1505\n",
+      "Epoch 64/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 64/100, Loss: 0.1466\n",
+      "Epoch 65/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 65/100, Loss: 0.1463\n",
+      "Epoch 66/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 66/100, Loss: 0.1517\n",
+      "Epoch 67/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 67/100, Loss: 0.1461\n",
+      "Epoch 68/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 68/100, Loss: 0.1428\n",
+      "Epoch 69/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 69/100, Loss: 0.1457\n",
+      "Epoch 70/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 70/100, Loss: 0.1549\n",
+      "Epoch 71/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 71/100, Loss: 0.1504\n",
+      "Epoch 72/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 72/100, Loss: 0.1455\n",
+      "Epoch 73/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 73/100, Loss: 0.1425\n",
+      "Epoch 74/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 74/100, Loss: 0.1422\n",
+      "Epoch 75/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 75/100, Loss: 0.1421\n",
+      "Epoch 76/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 76/100, Loss: 0.1452\n",
+      "Epoch 77/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 77/100, Loss: 0.1479\n",
+      "Epoch 78/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 78/100, Loss: 0.1371\n",
+      "Epoch 79/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 79/100, Loss: 0.1323\n",
+      "Epoch 80/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 80/100, Loss: 0.1396\n",
+      "Epoch 81/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 81/100, Loss: 0.1373\n",
+      "Epoch 82/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 82/100, Loss: 0.1366\n",
+      "Epoch 83/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 83/100, Loss: 0.1334\n",
+      "Epoch 84/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Epoch Progress:  84%|████████▍ | 101/120 [00:01<00:00, 70.34it/s, Loss=0.133]"
+     ]
+    }
+   ],
+   "source": [
+    "VAE_train(vae, optimizer, loader, epochs, dim_input, save_path=None)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 278,
+   "id": "3b7f5142-0db4-4ff9-8379-2a1d3db79537",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "f_test, ferr_test, specz_test = photoz_archive.get_testing_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 279,
+   "id": "0ea2e51f-1879-485e-a0c5-36b564ce5bc2",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "Ntest=10"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 301,
+   "id": "f2a2896d-cfa1-4978-a403-5a1ac62ddbfc",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "datain = torch.randn(size=(1, 50)).to(device)\n",
+    "x = vae.encoder(datain)\n",
+    "mu = vae.fc_mu(x)\n",
+    "log_var = vae.fc_logvar(x)\n",
+    "Nsamp=1000"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 303,
+   "id": "e1ee1209-7f4b-40e1-b7b5-a83c3fe8ef75",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "ppz = np.zeros(shape=(Ntest,Nsamp))\n",
+    "for ii in range(Ntest):\n",
+    "    for jj in range(Nsamp):\n",
+    "        z =vae.sampling(mu,log_var)\n",
+    "        ypred = vae.decode(z.to(device),torch.Tensor(f[ii]).unsqueeze(0).to(device))\n",
+    "        ppz[ii,jj] = ypred"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 304,
+   "id": "fec26b65-cdc3-4b11-8fd9-cd502cacfc81",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "m=9"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 305,
+   "id": "a1e92836-3465-44a1-b554-0de380e5ba16",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(array([ 3.,  2.,  3.,  7.,  4., 12., 12.,  9., 13., 16., 14., 23., 18.,\n",
+       "        25., 42., 56., 45., 49., 48., 34., 55., 42., 45., 39., 34., 41.,\n",
+       "        25., 33., 37., 26., 24., 21., 22., 27., 18., 19., 12., 12.,  4.,\n",
+       "         8.,  6.,  3.,  6.,  2.,  2.,  0.,  1.,  0.,  0.,  1.]),\n",
+       " array([0.56738353, 0.57165519, 0.57592686, 0.58019852, 0.58447019,\n",
+       "        0.58874185, 0.59301352, 0.59728518, 0.60155684, 0.60582851,\n",
+       "        0.61010017, 0.61437184, 0.6186435 , 0.62291517, 0.62718683,\n",
+       "        0.6314585 , 0.63573016, 0.64000183, 0.64427349, 0.64854516,\n",
+       "        0.65281682, 0.65708848, 0.66136015, 0.66563181, 0.66990348,\n",
+       "        0.67417514, 0.67844681, 0.68271847, 0.68699014, 0.6912618 ,\n",
+       "        0.69553347, 0.69980513, 0.7040768 , 0.70834846, 0.71262012,\n",
+       "        0.71689179, 0.72116345, 0.72543512, 0.72970678, 0.73397845,\n",
+       "        0.73825011, 0.74252178, 0.74679344, 0.75106511, 0.75533677,\n",
+       "        0.75960844, 0.7638801 , 0.76815176, 0.77242343, 0.77669509,\n",
+       "        0.78096676]),\n",
+       " <BarContainer object of 50 artists>)"
+      ]
+     },
+     "execution_count": 305,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.hist(ppz[m], bins =50)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 306,
+   "id": "a9f7342a-4b24-48bf-b4dc-3b74380fa042",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0.6869"
+      ]
+     },
+     "execution_count": 306,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "specz[m]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 284,
+   "id": "b0d9577b-5534-49be-8510-2e2ee65d7dce",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "OVERFITTING? DIFFERENCE TRAIN TEST? CHECK!"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 259,
+   "id": "21e8e786-5cf9-44f4-b6a7-e8fbb0d36d44",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 266,
+   "id": "49174198-d5e3-490d-b448-e509f07ac30f",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([1.0001, 1.0000, 1.0001, 1.0001, 0.9999, 1.0000, 1.0000, 1.0001, 0.9999,\n",
+       "        1.0000], device='cuda:0', grad_fn=<ExpBackward0>)"
+      ]
+     },
+     "execution_count": 266,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.exp(log_var[0])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 267,
+   "id": "f1248322-b424-4855-bc8b-ce8af1fcb275",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([-1.7762e-05, -3.3602e-06, -8.1182e-05, -1.8381e-05,  8.2459e-05,\n",
+       "         2.9923e-06,  1.5706e-04,  2.2795e-04, -1.3318e-05, -1.1017e-05],\n",
+       "       device='cuda:0', grad_fn=<SelectBackward0>)"
+      ]
+     },
+     "execution_count": 267,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "mu[0]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d14b8ae2-6bf7-47da-9725-57cc3f2b6cca",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "DLenv2",
+   "language": "python",
+   "name": "dlenv2"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.7"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

notebooks/.ipynb_checkpoints/CVAE-checkpoint.ipynb

@@ -0,0 +1,1833 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 84,
+   "id": "10a00e46-827a-4278-a715-99526591a0a7",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import torch.nn as nn\n",
+    "import torch\n",
+    "class CondVAE(nn.Module):\n",
+    "    def __init__(self, dim_input, latent_dim=10, context_vector_size=6):\n",
+    "        super(CondVAE, self).__init__()\n",
+    "        \n",
+    "        self.latent_dim = latent_dim\n",
+    "\n",
+    "        # Encoder\n",
+    "        self.encoder = nn.Sequential(\n",
+    "            nn.Linear(in_features=dim_input, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=300),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=300, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Flatten()\n",
+    "        )\n",
+    "        \n",
+    "        self.fc_mu = nn.Linear(100, latent_dim)\n",
+    "        self.fc_logvar = nn.Linear(100, latent_dim)\n",
+    "\n",
+    "        # Decoder\n",
+    "        self.decoder = nn.Sequential(\n",
+    "            nn.Linear(in_features=latent_dim+6, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=300),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=300, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=1),\n",
+    "        )\n",
+    "\n",
+    "    def encode(self, x):\n",
+    "        x = self.encoder(x)\n",
+    "        mu = self.fc_mu(x)\n",
+    "        log_var = self.fc_logvar(x)\n",
+    "\n",
+    "        return mu, log_var\n",
+    "    \n",
+    "    def decode(self, z, context_vector):\n",
+    "        # Concatenate the sampling (latent distribution) + embedding -> samples conditioned on both the input data and the specified label\n",
+    "        #print(properties.shape, z.shape)\n",
+    "        zcomb = torch.concat((z, context_vector), 1)\n",
+    "        #print(zcomb.shape)\n",
+    "        \n",
+    "        return self.decoder(zcomb)     \n",
+    "    \n",
+    "    def sampling(self, mu, log_var):\n",
+    "        # calculate standard deviation\n",
+    "        std = log_var.mul(0.5).exp_()\n",
+    "        \n",
+    "        # create noise tensor of same size as std to add to the latent vector\n",
+    "        eps = torch.cuda.FloatTensor(std.size()).normal_()\n",
+    "        \n",
+    "        # multiply eps with std to scale the random noise according to the learned distribution + add combined\n",
+    "        return eps.mul(std).add_(mu) # return z sample \n",
+    "\n",
+    "    def forward(self, x, context_vector):\n",
+    "        mu, log_var = self.encode(x)\n",
+    "        z = self.sampling(mu, log_var)\n",
+    "        #print(z.shape)\n",
+    "\n",
+    "        return self.decode(z, context_vector), mu, log_var\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 268,
+   "id": "8bbbe719-f9a3-4ace-ac4c-ed29ec4f9486",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import tqdm\n",
+    "import torch\n",
+    "import torch.nn as nn\n",
+    "\n",
+    "def condvae_loss(pred, label, mu, logvar):\n",
+    "    \"\"\"\n",
+    "    Calculate the conditional Variational Autoencoder (cVAE) loss.\n",
+    "\n",
+    "    This function computes the cVAE loss, which consists of two components:\n",
+    "    - Reconstruction loss: Measures the discrepancy between the reconstructed\n",
+    "      data and the original input.\n",
+    "    - KL divergence loss: Quantifies the difference between the learned latent\n",
+    "      distribution and the desired prior distribution (Gaussian).\n",
+    "\n",
+    "    Args:\n",
+    "        recon_x (torch.Tensor): Reconstructed data from the VAE.\n",
+    "        x (torch.Tensor): Original input data.\n",
+    "        mu (torch.Tensor): Latent variable mean.\n",
+    "        logvar (torch.Tensor): Logarithm of latent variable variance.\n",
+    "\n",
+    "    Returns:\n",
+    "        torch.Tensor: Computed cVAE loss.\n",
+    "    \"\"\"\n",
+    "    \n",
+    "    # MSE loss element-wise and sums up the individual losses\n",
+    "    cde_loss = nn.MSELoss(reduction='mean')(pred, label)\n",
+    "    \n",
+    "    # quantifies the difference between the learned latent distribution and the desired prior distribution (Gaussian)\n",
+    "    kl_divergence = -0.5 * torch.sum(1 + logvar - mu.pow(2) - logvar.exp())\n",
+    "    \n",
+    "    return cde_loss + kl_divergence\n",
+    "\n",
+    "def VAE_trainEpoch(model, optimizer, train_loader, dim_in=100):\n",
+    "    \"\"\"\n",
+    "    Train a Variational Autoencoder (VAE) for one epoch.\n",
+    "\n",
+    "    This function trains a VAE for one epoch using the provided data loader.\n",
+    "    It calculates the cVAE loss, performs backpropagation, and updates the model's parameters.\n",
+    "\n",
+    "    Args:\n",
+    "        model (nn.Module): VAE model to be trained.\n",
+    "        optimizer (torch.optim.Optimizer): Optimizer for updating model parameters.\n",
+    "        train_loader (DataLoader): DataLoader containing training data.\n",
+    "        dim_in (int): Dimensionality of the input noise.\n",
+    "\n",
+    "    Returns:\n",
+    "        float: Average loss for the epoch.\n",
+    "    \"\"\"\n",
+    "    model = model.train()\n",
+    "    device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
+    "    total_loss = 0\n",
+    "\n",
+    "    progress_bar = tqdm.tqdm(train_loader, desc=\"Epoch Progress\", leave=False)\n",
+    "    for context_vector, label in progress_bar:\n",
+    "        context_vector = context_vector.to(device)\n",
+    "        datain = torch.randn(size=(len(context_vector), dim_in)).to(device)\n",
+    "        label = label.unsqueeze(1).cuda()\n",
+    "        optimizer.zero_grad()\n",
+    "\n",
+    "        recon_batch, mu, log_var = model(datain, context_vector)\n",
+    "        loss = condvae_loss(recon_batch, label, mu, log_var)\n",
+    "\n",
+    "        loss.backward()\n",
+    "        optimizer.step()\n",
+    "\n",
+    "        total_loss += loss.item()\n",
+    "        progress_bar.set_postfix({\"Loss\": total_loss / (progress_bar.n + 1)})\n",
+    "\n",
+    "    return total_loss / len(train_loader)\n",
+    "\n",
+    "def VAE_train(model, optimizer, train_loader, epochs, dim_in, save_path=None):\n",
+    "    \"\"\"\n",
+    "    Train a Variational Autoencoder (VAE) for multiple epochs.\n",
+    "\n",
+    "    This function trains a VAE for the specified number of epochs using the provided data loader.\n",
+    "    It prints the epoch progress and the computed loss for each epoch.\n",
+    "\n",
+    "    Args:\n",
+    "        model (nn.Module): VAE model to be trained.\n",
+    "        optimizer (torch.optim.Optimizer): Optimizer for updating model parameters.\n",
+    "        train_loader (DataLoader): DataLoader containing training data.\n",
+    "        epochs (int): Number of epochs for training.\n",
+    "\n",
+    "    Returns:\n",
+    "        None\n",
+    "    \"\"\"\n",
+    "    for epoch in range(epochs):\n",
+    "        print(f\"Epoch {epoch + 1}/{epochs}\")\n",
+    "        epoch_loss = VAE_trainEpoch(model, optimizer, train_loader, dim_in)\n",
+    "        print(f\"Epoch {epoch + 1}/{epochs}, Loss: {epoch_loss:.4f}\")\n",
+    "        \n",
+    "    if save_path!=None:\n",
+    "        torch.save(model, save_path)\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 269,
+   "id": "11aaa0e9-e745-483a-887d-d851e791f8e4",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "from astropy.io import fits\n",
+    "import os\n",
+    "from astropy.table import Table\n",
+    "from scipy.spatial import KDTree\n",
+    "\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "from IPython.display import Image\n",
+    "from IPython.core.display import HTML "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 270,
+   "id": "0814440a-e341-4540-bfba-466b74b9873d",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import torch\n",
+    "from torch.utils.data import DataLoader, dataset, TensorDataset\n",
+    "from torch import nn, optim\n",
+    "from torch.optim import lr_scheduler"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 271,
+   "id": "8ecf9b60-bd03-4fa6-9516-c767d04b2071",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import sys\n",
+    "sys.path.append('../insight')\n",
+    "from archive import archive \n",
+    "from insight_arch import Photoz_network\n",
+    "from insight import Insight_module\n",
+    "from utils import sigma68, nmad, plot_photoz_estimates\n",
+    "from scipy import stats"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 272,
+   "id": "1277097d-e4bb-4bdd-b1f4-bccc88be0169",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "from matplotlib import rcParams\n",
+    "rcParams[\"mathtext.fontset\"] = \"stix\"\n",
+    "rcParams[\"font.family\"] = \"STIXGeneral\"\n",
+    "parent_dir = '/data/astro/scratch/lcabayol/Euclid/NNphotozs/Euclid_EXT_MER_PHZ_DC2_v1.5'"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 273,
+   "id": "5d2d3713-ff7f-4f16-860f-cf5ff42801b1",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "photoz_archive = archive(path = parent_dir, Qz_cut=1)\n",
+    "f, ferr, specz, specqz = photoz_archive.get_training_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 274,
+   "id": "45af2d9e-1160-4859-9888-f5daf62df84a",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "dset = TensorDataset(torch.Tensor(f),torch.Tensor(specz))\n",
+    "loader = DataLoader(dset, batch_size=100, shuffle=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 275,
+   "id": "dad8733c-c36a-4e86-b32d-41c244ba6259",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "dim_input=50\n",
+    "latent_dim=10\n",
+    "context_vector_dim=6\n",
+    "epochs=100\n",
+    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 276,
+   "id": "ef0b6560-dd3d-4b9d-a14c-43bf9338f7a4",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "vae = CondVAE(dim_input, latent_dim=latent_dim, context_vector_size=6).to(device)\n",
+    "optimizer = optim.Adam(vae.parameters(), lr=1e-3, weight_decay=1e-4)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "dd04d484-d14d-488d-a640-180fb6ab1001",
+   "metadata": {
+    "collapsed": true,
+    "jupyter": {
+     "outputs_hidden": true
+    },
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 1/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                            \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 1/100, Loss: 70.1670\n",
+      "Epoch 2/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                            \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 2/100, Loss: 5.8741\n",
+      "Epoch 3/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 3/100, Loss: 0.4596\n",
+      "Epoch 4/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 4/100, Loss: 0.9228\n",
+      "Epoch 5/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 5/100, Loss: 0.5939\n",
+      "Epoch 6/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 6/100, Loss: 0.7836\n",
+      "Epoch 7/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 7/100, Loss: 0.4829\n",
+      "Epoch 8/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 8/100, Loss: 0.5570\n",
+      "Epoch 9/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 9/100, Loss: 0.4176\n",
+      "Epoch 10/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                            \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 10/100, Loss: 0.9913\n",
+      "Epoch 11/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 11/100, Loss: 0.3367\n",
+      "Epoch 12/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 12/100, Loss: 0.3655\n",
+      "Epoch 13/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 13/100, Loss: 0.2642\n",
+      "Epoch 14/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 14/100, Loss: 0.2729\n",
+      "Epoch 15/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 15/100, Loss: 0.2490\n",
+      "Epoch 16/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 16/100, Loss: 0.2426\n",
+      "Epoch 17/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 17/100, Loss: 0.2418\n",
+      "Epoch 18/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 18/100, Loss: 0.2341\n",
+      "Epoch 19/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 19/100, Loss: 0.2288\n",
+      "Epoch 20/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 20/100, Loss: 0.2216\n",
+      "Epoch 21/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 21/100, Loss: 0.2200\n",
+      "Epoch 22/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 22/100, Loss: 0.2140\n",
+      "Epoch 23/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 23/100, Loss: 0.2125\n",
+      "Epoch 24/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 24/100, Loss: 0.2178\n",
+      "Epoch 25/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 25/100, Loss: 0.2127\n",
+      "Epoch 26/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 26/100, Loss: 0.2057\n",
+      "Epoch 27/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 27/100, Loss: 0.2168\n",
+      "Epoch 28/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 28/100, Loss: 0.2043\n",
+      "Epoch 29/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 29/100, Loss: 0.2039\n",
+      "Epoch 30/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 30/100, Loss: 0.1975\n",
+      "Epoch 31/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 31/100, Loss: 0.1956\n",
+      "Epoch 32/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 32/100, Loss: 0.1958\n",
+      "Epoch 33/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 33/100, Loss: 0.1893\n",
+      "Epoch 34/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 34/100, Loss: 0.1890\n",
+      "Epoch 35/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 35/100, Loss: 0.1869\n",
+      "Epoch 36/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 36/100, Loss: 0.1818\n",
+      "Epoch 37/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 37/100, Loss: 0.1784\n",
+      "Epoch 38/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 38/100, Loss: 0.1761\n",
+      "Epoch 39/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 39/100, Loss: 0.1757\n",
+      "Epoch 40/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 40/100, Loss: 0.1746\n",
+      "Epoch 41/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 41/100, Loss: 0.1756\n",
+      "Epoch 42/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 42/100, Loss: 0.1711\n",
+      "Epoch 43/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 43/100, Loss: 0.1706\n",
+      "Epoch 44/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 44/100, Loss: 0.1681\n",
+      "Epoch 45/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 45/100, Loss: 0.1672\n",
+      "Epoch 46/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 46/100, Loss: 0.1617\n",
+      "Epoch 47/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 47/100, Loss: 0.1637\n",
+      "Epoch 48/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 48/100, Loss: 0.1681\n",
+      "Epoch 49/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 49/100, Loss: 0.1637\n",
+      "Epoch 50/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 50/100, Loss: 0.1584\n",
+      "Epoch 51/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 51/100, Loss: 0.1591\n",
+      "Epoch 52/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 52/100, Loss: 0.1583\n",
+      "Epoch 53/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 53/100, Loss: 0.1536\n",
+      "Epoch 54/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 54/100, Loss: 0.1584\n",
+      "Epoch 55/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 55/100, Loss: 0.1624\n",
+      "Epoch 56/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 56/100, Loss: 0.1602\n",
+      "Epoch 57/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 57/100, Loss: 0.1547\n",
+      "Epoch 58/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 58/100, Loss: 0.1540\n",
+      "Epoch 59/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 59/100, Loss: 0.1541\n",
+      "Epoch 60/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 60/100, Loss: 0.1505\n",
+      "Epoch 61/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 61/100, Loss: 0.1521\n",
+      "Epoch 62/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 62/100, Loss: 0.1504\n",
+      "Epoch 63/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 63/100, Loss: 0.1505\n",
+      "Epoch 64/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 64/100, Loss: 0.1466\n",
+      "Epoch 65/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 65/100, Loss: 0.1463\n",
+      "Epoch 66/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 66/100, Loss: 0.1517\n",
+      "Epoch 67/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 67/100, Loss: 0.1461\n",
+      "Epoch 68/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 68/100, Loss: 0.1428\n",
+      "Epoch 69/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 69/100, Loss: 0.1457\n",
+      "Epoch 70/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 70/100, Loss: 0.1549\n",
+      "Epoch 71/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 71/100, Loss: 0.1504\n",
+      "Epoch 72/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 72/100, Loss: 0.1455\n",
+      "Epoch 73/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 73/100, Loss: 0.1425\n",
+      "Epoch 74/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 74/100, Loss: 0.1422\n",
+      "Epoch 75/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 75/100, Loss: 0.1421\n",
+      "Epoch 76/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 76/100, Loss: 0.1452\n",
+      "Epoch 77/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 77/100, Loss: 0.1479\n",
+      "Epoch 78/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 78/100, Loss: 0.1371\n",
+      "Epoch 79/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 79/100, Loss: 0.1323\n",
+      "Epoch 80/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 80/100, Loss: 0.1396\n",
+      "Epoch 81/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 81/100, Loss: 0.1373\n",
+      "Epoch 82/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 82/100, Loss: 0.1366\n",
+      "Epoch 83/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 83/100, Loss: 0.1334\n",
+      "Epoch 84/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Epoch Progress:  84%|████████▍ | 101/120 [00:01<00:00, 70.34it/s, Loss=0.133]"
+     ]
+    }
+   ],
+   "source": [
+    "VAE_train(vae, optimizer, loader, epochs, dim_input, save_path=None)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 278,
+   "id": "3b7f5142-0db4-4ff9-8379-2a1d3db79537",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "f_test, ferr_test, specz_test = photoz_archive.get_testing_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 279,
+   "id": "0ea2e51f-1879-485e-a0c5-36b564ce5bc2",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "Ntest=10"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 301,
+   "id": "f2a2896d-cfa1-4978-a403-5a1ac62ddbfc",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "datain = torch.randn(size=(1, 50)).to(device)\n",
+    "x = vae.encoder(datain)\n",
+    "mu = vae.fc_mu(x)\n",
+    "log_var = vae.fc_logvar(x)\n",
+    "Nsamp=1000"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 303,
+   "id": "e1ee1209-7f4b-40e1-b7b5-a83c3fe8ef75",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "ppz = np.zeros(shape=(Ntest,Nsamp))\n",
+    "for ii in range(Ntest):\n",
+    "    for jj in range(Nsamp):\n",
+    "        z =vae.sampling(mu,log_var)\n",
+    "        ypred = vae.decode(z.to(device),torch.Tensor(f[ii]).unsqueeze(0).to(device))\n",
+    "        ppz[ii,jj] = ypred"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 304,
+   "id": "fec26b65-cdc3-4b11-8fd9-cd502cacfc81",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "m=9"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 305,
+   "id": "a1e92836-3465-44a1-b554-0de380e5ba16",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(array([ 3.,  2.,  3.,  7.,  4., 12., 12.,  9., 13., 16., 14., 23., 18.,\n",
+       "        25., 42., 56., 45., 49., 48., 34., 55., 42., 45., 39., 34., 41.,\n",
+       "        25., 33., 37., 26., 24., 21., 22., 27., 18., 19., 12., 12.,  4.,\n",
+       "         8.,  6.,  3.,  6.,  2.,  2.,  0.,  1.,  0.,  0.,  1.]),\n",
+       " array([0.56738353, 0.57165519, 0.57592686, 0.58019852, 0.58447019,\n",
+       "        0.58874185, 0.59301352, 0.59728518, 0.60155684, 0.60582851,\n",
+       "        0.61010017, 0.61437184, 0.6186435 , 0.62291517, 0.62718683,\n",
+       "        0.6314585 , 0.63573016, 0.64000183, 0.64427349, 0.64854516,\n",
+       "        0.65281682, 0.65708848, 0.66136015, 0.66563181, 0.66990348,\n",
+       "        0.67417514, 0.67844681, 0.68271847, 0.68699014, 0.6912618 ,\n",
+       "        0.69553347, 0.69980513, 0.7040768 , 0.70834846, 0.71262012,\n",
+       "        0.71689179, 0.72116345, 0.72543512, 0.72970678, 0.73397845,\n",
+       "        0.73825011, 0.74252178, 0.74679344, 0.75106511, 0.75533677,\n",
+       "        0.75960844, 0.7638801 , 0.76815176, 0.77242343, 0.77669509,\n",
+       "        0.78096676]),\n",
+       " <BarContainer object of 50 artists>)"
+      ]
+     },
+     "execution_count": 305,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.hist(ppz[m], bins =50)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 306,
+   "id": "a9f7342a-4b24-48bf-b4dc-3b74380fa042",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0.6869"
+      ]
+     },
+     "execution_count": 306,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "specz[m]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 284,
+   "id": "b0d9577b-5534-49be-8510-2e2ee65d7dce",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "OVERFITTING? DIFFERENCE TRAIN TEST? CHECK!"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 259,
+   "id": "21e8e786-5cf9-44f4-b6a7-e8fbb0d36d44",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 266,
+   "id": "49174198-d5e3-490d-b448-e509f07ac30f",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([1.0001, 1.0000, 1.0001, 1.0001, 0.9999, 1.0000, 1.0000, 1.0001, 0.9999,\n",
+       "        1.0000], device='cuda:0', grad_fn=<ExpBackward0>)"
+      ]
+     },
+     "execution_count": 266,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.exp(log_var[0])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 267,
+   "id": "f1248322-b424-4855-bc8b-ce8af1fcb275",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([-1.7762e-05, -3.3602e-06, -8.1182e-05, -1.8381e-05,  8.2459e-05,\n",
+       "         2.9923e-06,  1.5706e-04,  2.2795e-04, -1.3318e-05, -1.1017e-05],\n",
+       "       device='cuda:0', grad_fn=<SelectBackward0>)"
+      ]
+     },
+     "execution_count": 267,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "mu[0]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d14b8ae2-6bf7-47da-9725-57cc3f2b6cca",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "DLenv2",
+   "language": "python",
+   "name": "dlenv2"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.7"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

notebooks/.ipynb_checkpoints/match_catalogues-checkpoint.ipynb

@@ -0,0 +1,6 @@
+{
+ "cells": [],
+ "metadata": {},
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

notebooks/.ipynb_checkpoints/toy_test-checkpoint.ipynb

@@ -0,0 +1,6 @@
+{
+ "cells": [],
+ "metadata": {},
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

notebooks/CVAE-Copy1.ipynb

@@ -0,0 +1,2177 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 293,
+   "id": "10a00e46-827a-4278-a715-99526591a0a7",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import torch.nn as nn\n",
+    "import torch\n",
+    "class CondVAE(nn.Module):\n",
+    "    def __init__(self, dim_input, latent_dim=10):\n",
+    "        super(CondVAE, self).__init__()\n",
+    "        \n",
+    "        self.latent_dim = latent_dim\n",
+    "\n",
+    "        # Encoder\n",
+    "        self.encoder = nn.Sequential(\n",
+    "            nn.Linear(in_features=dim_input, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=300),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=300, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Flatten()\n",
+    "        )\n",
+    "        \n",
+    "        self.fc_mu = nn.Linear(100, latent_dim)\n",
+    "        #self.fc_logvar =  nn.Sequential(nn.Linear(100, latent_dim),nn.Softplus())\n",
+    "        self.fc_logvar =  nn.Linear(100, latent_dim)\n",
+    "\n",
+    "        # Decoder\n",
+    "        self.decoder = nn.Sequential(\n",
+    "            nn.Linear(in_features=latent_dim, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=300),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=300, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=6),\n",
+    "        )\n",
+    "        \n",
+    "        self.regressor = nn.Sequential(\n",
+    "            nn.Linear(in_features=latent_dim, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=300),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=300, out_features=200),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=200, out_features=100),\n",
+    "            nn.ReLU(),\n",
+    "            nn.Linear(in_features=100, out_features=1),\n",
+    "        )\n",
+    "\n",
+    "    def encode(self, x):\n",
+    "        x = self.encoder(x)\n",
+    "        mu = self.fc_mu(x)\n",
+    "        log_var = self.fc_logvar(x)\n",
+    "\n",
+    "        return mu, log_var\n",
+    "    \n",
+    "    def decode(self, z):\n",
+    "        \n",
+    "        return self.decoder(z)     \n",
+    "    \n",
+    "    def sampling(self, mu, log_var):\n",
+    "        # calculate standard deviation\n",
+    "        std = log_var.mul(0.5).exp_()\n",
+    "        # create noise tensor of same size as std to add to the latent vector\n",
+    "        eps = torch.cuda.FloatTensor(std.size()).normal_()\n",
+    "        # multiply eps with std to scale the random noise according to the learned distribution + add combined\n",
+    "        return eps.mul(std).add_(mu) # return z sample \n",
+    "\n",
+    "    def forward(self, x):\n",
+    "        mu, log_var = self.encode(x)\n",
+    "        z = self.sampling(mu, log_var)\n",
+    "        \n",
+    "        return self.decode(z), self.regressor(z), mu, log_var\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 294,
+   "id": "8bbbe719-f9a3-4ace-ac4c-ed29ec4f9486",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import tqdm\n",
+    "import torch\n",
+    "import torch.nn as nn\n",
+    "\n",
+    "def condvae_loss(recons, pred, data_input, label, mu, logvar):\n",
+    "    \"\"\"\n",
+    "    Calculate the conditional Variational Autoencoder (cVAE) loss.\n",
+    "\n",
+    "    This function computes the cVAE loss, which consists of two components:\n",
+    "    - Reconstruction loss: Measures the discrepancy between the reconstructed\n",
+    "      data and the original input.\n",
+    "    - KL divergence loss: Quantifies the difference between the learned latent\n",
+    "      distribution and the desired prior distribution (Gaussian).\n",
+    "\n",
+    "    Args:\n",
+    "        recon_x (torch.Tensor): Reconstructed data from the VAE.\n",
+    "        x (torch.Tensor): Original input data.\n",
+    "        mu (torch.Tensor): Latent variable mean.\n",
+    "        logvar (torch.Tensor): Logarithm of latent variable variance.\n",
+    "\n",
+    "    Returns:\n",
+    "        torch.Tensor: Computed cVAE loss.\n",
+    "    \"\"\"\n",
+    "    \n",
+    "    # MSE loss element-wise and sums up the individual losses\n",
+    "    regression_loss = nn.L1Loss(reduction='mean')(pred, label)\n",
+    "    decoder_loss = nn.L1Loss(reduction='mean')(recons, data_input)\n",
+    "    \n",
+    "    # quantifies the difference between the learned latent distribution and the desired prior distribution (Gaussian)\n",
+    "    kl_divergence =  0 #-0.5 * torch.sum(1 + logvar - mu.pow(2) - logvar.exp())\n",
+    "    \n",
+    "    \n",
+    "    return regression_loss + kl_divergence + decoder_loss\n",
+    "\n",
+    "def VAE_trainEpoch(model, optimizer, train_loader, dim_in=100):\n",
+    "    \"\"\"\n",
+    "    Train a Variational Autoencoder (VAE) for one epoch.\n",
+    "\n",
+    "    This function trains a VAE for one epoch using the provided data loader.\n",
+    "    It calculates the cVAE loss, performs backpropagation, and updates the model's parameters.\n",
+    "\n",
+    "    Args:\n",
+    "        model (nn.Module): VAE model to be trained.\n",
+    "        optimizer (torch.optim.Optimizer): Optimizer for updating model parameters.\n",
+    "        train_loader (DataLoader): DataLoader containing training data.\n",
+    "        dim_in (int): Dimensionality of the input noise.\n",
+    "\n",
+    "    Returns:\n",
+    "        float: Average loss for the epoch.\n",
+    "    \"\"\"\n",
+    "    model = model.train()\n",
+    "    device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
+    "    total_loss = 0\n",
+    "\n",
+    "    progress_bar = tqdm.tqdm(train_loader, desc=\"Epoch Progress\", leave=False)\n",
+    "    for data, label in progress_bar:\n",
+    "        data = data.to(device)\n",
+    "        label = label.unsqueeze(1).cuda()\n",
+    "        optimizer.zero_grad()\n",
+    "\n",
+    "\n",
+    "        recon_batch, pred_batch, mu, log_var = model(data)\n",
+    "        loss = condvae_loss(recon_batch, pred_batch, data, label, mu, log_var)\n",
+    "\n",
+    "        loss.backward()\n",
+    "        optimizer.step()\n",
+    "\n",
+    "        total_loss += loss.item()\n",
+    "        progress_bar.set_postfix({\"Loss\": total_loss / (progress_bar.n + 1)})\n",
+    "\n",
+    "    return total_loss / len(train_loader)\n",
+    "\n",
+    "def VAE_train(model, optimizer, train_loader, epochs, dim_in, save_path=None):\n",
+    "    \"\"\"\n",
+    "    Train a Variational Autoencoder (VAE) for multiple epochs.\n",
+    "\n",
+    "    This function trains a VAE for the specified number of epochs using the provided data loader.\n",
+    "    It prints the epoch progress and the computed loss for each epoch.\n",
+    "\n",
+    "    Args:\n",
+    "        model (nn.Module): VAE model to be trained.\n",
+    "        optimizer (torch.optim.Optimizer): Optimizer for updating model parameters.\n",
+    "        train_loader (DataLoader): DataLoader containing training data.\n",
+    "        epochs (int): Number of epochs for training.\n",
+    "\n",
+    "    Returns:\n",
+    "        None\n",
+    "    \"\"\"\n",
+    "    for epoch in range(epochs):\n",
+    "        print(f\"Epoch {epoch + 1}/{epochs}\")\n",
+    "        epoch_loss = VAE_trainEpoch(model, optimizer, train_loader, dim_in)\n",
+    "        print(f\"Epoch {epoch + 1}/{epochs}, Loss: {epoch_loss:.4f}\")\n",
+    "        \n",
+    "    if save_path!=None:\n",
+    "        torch.save(model, save_path)\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 295,
+   "id": "11aaa0e9-e745-483a-887d-d851e791f8e4",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "from astropy.io import fits\n",
+    "import os\n",
+    "from astropy.table import Table\n",
+    "from scipy.spatial import KDTree\n",
+    "\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "from IPython.display import Image\n",
+    "from IPython.core.display import HTML "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 296,
+   "id": "0814440a-e341-4540-bfba-466b74b9873d",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import torch\n",
+    "from torch.utils.data import DataLoader, dataset, TensorDataset\n",
+    "from torch import nn, optim\n",
+    "from torch.optim import lr_scheduler"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 297,
+   "id": "8ecf9b60-bd03-4fa6-9516-c767d04b2071",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import sys\n",
+    "sys.path.append('../insight')\n",
+    "from archive import archive \n",
+    "from insight_arch import Photoz_network\n",
+    "from insight import Insight_module\n",
+    "from utils import sigma68, nmad, plot_photoz_estimates\n",
+    "from scipy import stats"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 298,
+   "id": "1277097d-e4bb-4bdd-b1f4-bccc88be0169",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "from matplotlib import rcParams\n",
+    "rcParams[\"mathtext.fontset\"] = \"stix\"\n",
+    "rcParams[\"font.family\"] = \"STIXGeneral\"\n",
+    "parent_dir = '/data/astro/scratch/lcabayol/Euclid/NNphotozs/Euclid_EXT_MER_PHZ_DC2_v1.5'"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 65,
+   "id": "5d2d3713-ff7f-4f16-860f-cf5ff42801b1",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "photoz_archive = archive(path = parent_dir, Qz_cut=1)\n",
+    "f, ferr, specz, specqz = photoz_archive.get_training_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 299,
+   "id": "45af2d9e-1160-4859-9888-f5daf62df84a",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "dset = TensorDataset(torch.Tensor(f),torch.Tensor(specz))\n",
+    "loader = DataLoader(dset, batch_size=100, shuffle=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 300,
+   "id": "dad8733c-c36a-4e86-b32d-41c244ba6259",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "dim_input=6\n",
+    "latent_dim=1\n",
+    "epochs=100\n",
+    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 301,
+   "id": "ef0b6560-dd3d-4b9d-a14c-43bf9338f7a4",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "vae = CondVAE(dim_input, latent_dim=latent_dim).to(device)\n",
+    "optimizer = optim.Adam(vae.parameters(), lr=1e-3, weight_decay=1e-4)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 302,
+   "id": "dd04d484-d14d-488d-a640-180fb6ab1001",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 1/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 1/100, Loss: 0.7994\n",
+      "Epoch 2/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 2/100, Loss: 0.7418\n",
+      "Epoch 3/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 3/100, Loss: 0.7436\n",
+      "Epoch 4/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 4/100, Loss: 0.7406\n",
+      "Epoch 5/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 5/100, Loss: 0.7395\n",
+      "Epoch 6/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 6/100, Loss: 0.7416\n",
+      "Epoch 7/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 7/100, Loss: 0.7402\n",
+      "Epoch 8/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 8/100, Loss: 0.7400\n",
+      "Epoch 9/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 9/100, Loss: 0.7397\n",
+      "Epoch 10/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 10/100, Loss: 0.7380\n",
+      "Epoch 11/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 11/100, Loss: 0.7407\n",
+      "Epoch 12/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 12/100, Loss: 0.7381\n",
+      "Epoch 13/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 13/100, Loss: 0.7408\n",
+      "Epoch 14/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 14/100, Loss: 0.7390\n",
+      "Epoch 15/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 15/100, Loss: 0.7387\n",
+      "Epoch 16/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 16/100, Loss: 0.7384\n",
+      "Epoch 17/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 17/100, Loss: 0.7382\n",
+      "Epoch 18/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 18/100, Loss: 0.7377\n",
+      "Epoch 19/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 19/100, Loss: 0.7382\n",
+      "Epoch 20/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 20/100, Loss: 0.7383\n",
+      "Epoch 21/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 21/100, Loss: 0.7385\n",
+      "Epoch 22/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 22/100, Loss: 0.7376\n",
+      "Epoch 23/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 23/100, Loss: 0.7372\n",
+      "Epoch 24/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 24/100, Loss: 0.7380\n",
+      "Epoch 25/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 25/100, Loss: 0.7373\n",
+      "Epoch 26/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 26/100, Loss: 0.7374\n",
+      "Epoch 27/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 27/100, Loss: 0.7373\n",
+      "Epoch 28/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 28/100, Loss: 0.7367\n",
+      "Epoch 29/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 29/100, Loss: 0.7371\n",
+      "Epoch 30/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 30/100, Loss: 0.7370\n",
+      "Epoch 31/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 31/100, Loss: 0.7377\n",
+      "Epoch 32/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 32/100, Loss: 0.7372\n",
+      "Epoch 33/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 33/100, Loss: 0.7365\n",
+      "Epoch 34/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 34/100, Loss: 0.7374\n",
+      "Epoch 35/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 35/100, Loss: 0.7362\n",
+      "Epoch 36/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 36/100, Loss: 0.7370\n",
+      "Epoch 37/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 37/100, Loss: 0.7367\n",
+      "Epoch 38/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 38/100, Loss: 0.7370\n",
+      "Epoch 39/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 39/100, Loss: 0.7363\n",
+      "Epoch 40/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 40/100, Loss: 0.7367\n",
+      "Epoch 41/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 41/100, Loss: 0.7368\n",
+      "Epoch 42/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 42/100, Loss: 0.7376\n",
+      "Epoch 43/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 43/100, Loss: 0.7364\n",
+      "Epoch 44/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 44/100, Loss: 0.7368\n",
+      "Epoch 45/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 45/100, Loss: 0.7366\n",
+      "Epoch 46/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 46/100, Loss: 0.7366\n",
+      "Epoch 47/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 47/100, Loss: 0.7371\n",
+      "Epoch 48/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 48/100, Loss: 0.7367\n",
+      "Epoch 49/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 49/100, Loss: 0.7362\n",
+      "Epoch 50/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 50/100, Loss: 0.7362\n",
+      "Epoch 51/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 51/100, Loss: 0.7366\n",
+      "Epoch 52/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 52/100, Loss: 0.7361\n",
+      "Epoch 53/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 53/100, Loss: 0.7364\n",
+      "Epoch 54/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 54/100, Loss: 0.7363\n",
+      "Epoch 55/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 55/100, Loss: 0.7362\n",
+      "Epoch 56/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 56/100, Loss: 0.7369\n",
+      "Epoch 57/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 57/100, Loss: 0.7366\n",
+      "Epoch 58/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 58/100, Loss: 0.7361\n",
+      "Epoch 59/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 59/100, Loss: 0.7364\n",
+      "Epoch 60/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 60/100, Loss: 0.7368\n",
+      "Epoch 61/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 61/100, Loss: 0.7362\n",
+      "Epoch 62/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 62/100, Loss: 0.7361\n",
+      "Epoch 63/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 63/100, Loss: 0.7375\n",
+      "Epoch 64/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 64/100, Loss: 0.7364\n",
+      "Epoch 65/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 65/100, Loss: 0.7369\n",
+      "Epoch 66/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 66/100, Loss: 0.7360\n",
+      "Epoch 67/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 67/100, Loss: 0.7361\n",
+      "Epoch 68/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 68/100, Loss: 0.7364\n",
+      "Epoch 69/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 69/100, Loss: 0.7363\n",
+      "Epoch 70/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 70/100, Loss: 0.7360\n",
+      "Epoch 71/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 71/100, Loss: 0.7364\n",
+      "Epoch 72/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 72/100, Loss: 0.7366\n",
+      "Epoch 73/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 73/100, Loss: 0.7364\n",
+      "Epoch 74/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 74/100, Loss: 0.7363\n",
+      "Epoch 75/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 75/100, Loss: 0.7360\n",
+      "Epoch 76/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 76/100, Loss: 0.7366\n",
+      "Epoch 77/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 77/100, Loss: 0.7364\n",
+      "Epoch 78/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 78/100, Loss: 0.7365\n",
+      "Epoch 79/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 79/100, Loss: 0.7365\n",
+      "Epoch 80/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 80/100, Loss: 0.7367\n",
+      "Epoch 81/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 81/100, Loss: 0.7368\n",
+      "Epoch 82/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 82/100, Loss: 0.7366\n",
+      "Epoch 83/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 83/100, Loss: 0.7360\n",
+      "Epoch 84/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 84/100, Loss: 0.7358\n",
+      "Epoch 85/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 85/100, Loss: 0.7368\n",
+      "Epoch 86/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 86/100, Loss: 0.7363\n",
+      "Epoch 87/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 87/100, Loss: 0.7359\n",
+      "Epoch 88/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 88/100, Loss: 0.7364\n",
+      "Epoch 89/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 89/100, Loss: 0.7361\n",
+      "Epoch 90/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 90/100, Loss: 0.7365\n",
+      "Epoch 91/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 91/100, Loss: 0.7361\n",
+      "Epoch 92/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 92/100, Loss: 0.7367\n",
+      "Epoch 93/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 93/100, Loss: 0.7362\n",
+      "Epoch 94/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 94/100, Loss: 0.7361\n",
+      "Epoch 95/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 95/100, Loss: 0.7362\n",
+      "Epoch 96/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 96/100, Loss: 0.7363\n",
+      "Epoch 97/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 97/100, Loss: 0.7363\n",
+      "Epoch 98/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 98/100, Loss: 0.7363\n",
+      "Epoch 99/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             \r"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 99/100, Loss: 0.7364\n",
+      "Epoch 100/100\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "                                                                             "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 100/100, Loss: 0.7364\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\r"
+     ]
+    }
+   ],
+   "source": [
+    "VAE_train(vae, optimizer, loader, epochs, dim_input, save_path=None)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 305,
+   "id": "3b7f5142-0db4-4ff9-8379-2a1d3db79537",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "f_test, ferr_test, specz_test = photoz_archive.get_testing_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 306,
+   "id": "0ea2e51f-1879-485e-a0c5-36b564ce5bc2",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "Ntest=10"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 307,
+   "id": "f2a2896d-cfa1-4978-a403-5a1ac62ddbfc",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "datain = torch.Tensor(f_test[:10]).to(device)\n",
+    "x = vae.encoder(datain)\n",
+    "mu = vae.fc_mu(x)\n",
+    "log_var = vae.fc_logvar(x)\n",
+    "Nsamp=1000"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 308,
+   "id": "9fa354f3-6d7f-441e-8c09-a815f2e212d5",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([[1.1479, 0.3912, 0.8374, 0.7508, 1.2488, 0.8488],\n",
+       "        [0.3916, 0.8170, 0.9124, 0.8022, 0.9451, 0.9885],\n",
+       "        [0.3894, 0.5246, 0.9165, 0.8517, 0.8625, 0.8356],\n",
+       "        [0.4891, 0.5986, 0.9832, 0.7717, 0.9078, 0.8872],\n",
+       "        [0.1693, 0.4140, 0.6905, 0.7985, 0.7315, 0.5996],\n",
+       "        [0.1097, 0.3383, 0.4823, 0.7898, 0.7083, 0.6996],\n",
+       "        [0.5493, 0.6261, 1.0609, 0.7438, 0.8732, 0.9139],\n",
+       "        [0.5659, 0.6637, 0.8864, 0.9856, 0.8920, 0.7977],\n",
+       "        [0.3160, 0.3441, 0.4192, 0.6507, 0.6579, 0.5727],\n",
+       "        [0.7710, 0.5087, 0.7844, 0.9328, 0.7800, 0.8684]], device='cuda:0')"
+      ]
+     },
+     "execution_count": 308,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "datain"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 309,
+   "id": "77ba620c-4146-4d55-94f2-645d2b7e19be",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([[-3.4548e-06],\n",
+       "        [-3.4548e-06],\n",
+       "        [-3.4548e-06],\n",
+       "        [-3.4548e-06],\n",
+       "        [-3.4548e-06],\n",
+       "        [-3.4548e-06],\n",
+       "        [-3.4548e-06],\n",
+       "        [-3.4548e-06],\n",
+       "        [-3.4548e-06],\n",
+       "        [-3.4548e-06]], device='cuda:0', grad_fn=<AddmmBackward0>)"
+      ]
+     },
+     "execution_count": 309,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "mu"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 310,
+   "id": "20871270-a53d-47e2-b3d8-606b486126b5",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([[2.0000],\n",
+       "        [2.0000],\n",
+       "        [2.0000],\n",
+       "        [2.0000],\n",
+       "        [2.0000],\n",
+       "        [2.0000],\n",
+       "        [2.0000],\n",
+       "        [2.0000],\n",
+       "        [2.0000],\n",
+       "        [2.0000]], device='cuda:0', grad_fn=<ExpBackward0>)"
+      ]
+     },
+     "execution_count": 310,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.exp(log_var)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 311,
+   "id": "47cd277a-e23a-401a-965a-dd7d93f7c196",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import torch.distributions as D\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 312,
+   "id": "d599b51a-f42a-4fc3-9acd-0902a985d4b1",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "tensor([2.0000], grad_fn=<ExpBackward0>)"
+      ]
+     },
+     "execution_count": 312,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.exp(log_var[ii].cpu())"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 313,
+   "id": "e1ee1209-7f4b-40e1-b7b5-a83c3fe8ef75",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "vae = vae.eval()\n",
+    "z_dim=1\n",
+    "py_z = np.zeros(shape=(Ntest,Nsamp))\n",
+    "px_z = np.zeros(shape=(Ntest,Nsamp,6))\n",
+    "pz= np.zeros(shape=(Ntest,Nsamp))\n",
+    "\n",
+    "for ii in range(Ntest):\n",
+    "    base_distribution = D.Normal(mu[ii].cpu()*torch.ones(z_dim), torch.exp(log_var[ii].cpu())*torch.ones(z_dim))\n",
+    "    for jj in range(Nsamp):\n",
+    "        z = vae.sampling(mu[ii],log_var[ii])        \n",
+    "\n",
+    "        py_z[ii,jj] = vae.regressor(z.to(device)).detach().cpu().numpy()\n",
+    "        px_z[ii,jj,:] = vae.decode(z.to(device)).detach().cpu().numpy()\n",
+    "        pz[ii,jj] = base_distribution.log_prob(z.cpu())\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 314,
+   "id": "258a6609-9a07-4313-8d7e-581bf623ce71",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([[0.12801583, 0.17218088, 0.1873554 , ..., 0.14763936, 0.15918671,\n",
+       "        0.10822631],\n",
+       "       [0.17516595, 0.19795613, 0.10892031, ..., 0.09822501, 0.19940973,\n",
+       "        0.15055025],\n",
+       "       [0.16263473, 0.17919608, 0.14378181, ..., 0.19080766, 0.19529714,\n",
+       "        0.18262688],\n",
+       "       ...,\n",
+       "       [0.0383369 , 0.1969846 , 0.19438929, ..., 0.15335193, 0.19859305,\n",
+       "        0.19859523],\n",
+       "       [0.18836775, 0.19947221, 0.15643367, ..., 0.19192475, 0.19785437,\n",
+       "        0.19929699],\n",
+       "       [0.19779257, 0.14646251, 0.19313204, ..., 0.19487876, 0.19082342,\n",
+       "        0.18486699]])"
+      ]
+     },
+     "execution_count": 314,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "np.exp(pz)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 316,
+   "id": "a1e92836-3465-44a1-b554-0de380e5ba16",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(array([   0.,    0.,    0.,    0.,    0.,    0.,    0.,    0.,    0.,\n",
+       "           0.,    0.,    0.,    0.,    0.,    0.,    0., 1000.,    0.,\n",
+       "           0.,    0.,    0.,    0.,    0.,    0.,    0.,    0.,    0.,\n",
+       "           0.,    0.,    0.,    0.,    0.,    0.,    0.,    0.,    0.,\n",
+       "           0.,    0.,    0.,    0.,    0.,    0.,    0.,    0.,    0.,\n",
+       "           0.,    0.,    0.,    0.,    0.]),\n",
+       " array([0.  , 0.04, 0.08, 0.12, 0.16, 0.2 , 0.24, 0.28, 0.32, 0.36, 0.4 ,\n",
+       "        0.44, 0.48, 0.52, 0.56, 0.6 , 0.64, 0.68, 0.72, 0.76, 0.8 , 0.84,\n",
+       "        0.88, 0.92, 0.96, 1.  , 1.04, 1.08, 1.12, 1.16, 1.2 , 1.24, 1.28,\n",
+       "        1.32, 1.36, 1.4 , 1.44, 1.48, 1.52, 1.56, 1.6 , 1.64, 1.68, 1.72,\n",
+       "        1.76, 1.8 , 1.84, 1.88, 1.92, 1.96, 2.  ]),\n",
+       " <BarContainer object of 50 artists>)"
+      ]
+     },
+     "execution_count": 316,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.hist(py_z[m], bins =50, range =(0,2))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 318,
+   "id": "d14b8ae2-6bf7-47da-9725-57cc3f2b6cca",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "1.103"
+      ]
+     },
+     "execution_count": 318,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "specz_test[m]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c2962376-46bc-4153-bfe5-219e10369709",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "DLenv2",
+   "language": "python",
+   "name": "dlenv2"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.7"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

notebooks/PLOTS.ipynb

@@ -0,0 +1,579 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "873187db-8223-4aa9-88ef-9ee3cc8fbfa4",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "from astropy.io import fits\n",
+    "import os\n",
+    "from astropy.table import Table\n",
+    "from scipy.spatial import KDTree\n",
+    "\n",
+    "import matplotlib.pyplot as plt\n",
+    "\n",
+    "from IPython.display import Image\n",
+    "from IPython.core.display import HTML "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "0cd7baac-bec6-4619-bf86-b03baf28ea8c",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/data/astro/scratch/lcabayol/anaconda3/envs/DESIenv6/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
+      "  from .autonotebook import tqdm as notebook_tqdm\n"
+     ]
+    }
+   ],
+   "source": [
+    "import torch\n",
+    "from torch.utils.data import DataLoader, dataset, TensorDataset\n",
+    "from torch import nn, optim\n",
+    "from torch.optim import lr_scheduler"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "8a694b63-85ec-49b9-9836-c5b579d94281",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import sys\n",
+    "sys.path.append('../insight')\n",
+    "from archive import archive \n",
+    "from insight_arch import Photoz_network\n",
+    "from insight import Insight_module\n",
+    "from utils import sigma68, nmad, plot_photoz\n",
+    "from scipy import stats"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "6f50d39b-eac8-4f49-a4b7-7579c9984a61",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from matplotlib import rcParams\n",
+    "rcParams[\"mathtext.fontset\"] = \"stix\"\n",
+    "rcParams[\"font.family\"] = \"STIXGeneral\"\n",
+    "parent_dir = '/data/astro/scratch/lcabayol/Euclid/NNphotozs/Euclid_EXT_MER_PHZ_DC2_v1.5'"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "661b9a50-684f-4e7d-9293-a73ec5edb98f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "photoz_archive = archive(path = parent_dir, Qz_cut=1)\n",
+    "f, ferr, specz, specqz = photoz_archive.get_training_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 100,
+   "id": "eff8e565-4e6e-41a0-ad54-4035bea6b14b",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "f_test, ferr_test, specz_test = photoz_archive.get_testing_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "667e4edc-1a58-438d-b1ef-e01ad199e79f",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 74,
+   "id": "ace74805-afb6-4d05-826a-901c0e115b8d",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "df_test = pd.read_csv('/data/astro/scratch/lcabayol/Euclid/NNphotozs/results/df1.csv', sep=',', header = 0, comment='#')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 76,
+   "id": "854d72a7-e748-4e79-8f18-0c80b2a58ed8",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>VISmag</th>\n",
+       "      <th>zs</th>\n",
+       "      <th>z</th>\n",
+       "      <th>zuncert</th>\n",
+       "      <th>zwerr</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>23.103798</td>\n",
+       "      <td>1.103000</td>\n",
+       "      <td>1.077487</td>\n",
+       "      <td>0.147231</td>\n",
+       "      <td>-0.012132</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>22.471019</td>\n",
+       "      <td>0.468800</td>\n",
+       "      <td>0.416247</td>\n",
+       "      <td>0.177303</td>\n",
+       "      <td>-0.035780</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>21.853940</td>\n",
+       "      <td>0.694600</td>\n",
+       "      <td>0.639316</td>\n",
+       "      <td>0.124684</td>\n",
+       "      <td>-0.032623</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>22.005561</td>\n",
+       "      <td>0.649200</td>\n",
+       "      <td>0.628935</td>\n",
+       "      <td>0.128350</td>\n",
+       "      <td>-0.012288</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>22.204387</td>\n",
+       "      <td>0.666900</td>\n",
+       "      <td>0.611376</td>\n",
+       "      <td>0.104980</td>\n",
+       "      <td>-0.033309</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12048</th>\n",
+       "      <td>22.449399</td>\n",
+       "      <td>0.690462</td>\n",
+       "      <td>0.722806</td>\n",
+       "      <td>0.123866</td>\n",
+       "      <td>0.019133</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12049</th>\n",
+       "      <td>22.102501</td>\n",
+       "      <td>0.915746</td>\n",
+       "      <td>0.956847</td>\n",
+       "      <td>0.117305</td>\n",
+       "      <td>0.021454</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12050</th>\n",
+       "      <td>22.982543</td>\n",
+       "      <td>0.721060</td>\n",
+       "      <td>0.745688</td>\n",
+       "      <td>0.180621</td>\n",
+       "      <td>0.014309</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12051</th>\n",
+       "      <td>20.037661</td>\n",
+       "      <td>0.345100</td>\n",
+       "      <td>0.358207</td>\n",
+       "      <td>0.070814</td>\n",
+       "      <td>0.009744</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12052</th>\n",
+       "      <td>22.764413</td>\n",
+       "      <td>0.487737</td>\n",
+       "      <td>0.363416</td>\n",
+       "      <td>0.228689</td>\n",
+       "      <td>-0.083564</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>12053 rows × 5 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "          VISmag        zs         z   zuncert     zwerr\n",
+       "0      23.103798  1.103000  1.077487  0.147231 -0.012132\n",
+       "1      22.471019  0.468800  0.416247  0.177303 -0.035780\n",
+       "2      21.853940  0.694600  0.639316  0.124684 -0.032623\n",
+       "3      22.005561  0.649200  0.628935  0.128350 -0.012288\n",
+       "4      22.204387  0.666900  0.611376  0.104980 -0.033309\n",
+       "...          ...       ...       ...       ...       ...\n",
+       "12048  22.449399  0.690462  0.722806  0.123866  0.019133\n",
+       "12049  22.102501  0.915746  0.956847  0.117305  0.021454\n",
+       "12050  22.982543  0.721060  0.745688  0.180621  0.014309\n",
+       "12051  20.037661  0.345100  0.358207  0.070814  0.009744\n",
+       "12052  22.764413  0.487737  0.363416  0.228689 -0.083564\n",
+       "\n",
+       "[12053 rows x 5 columns]"
+      ]
+     },
+     "execution_count": 76,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df_test"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d72d3057-b2d6-42ee-8729-0ea6f18c0028",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 77,
+   "id": "78f955d8-8165-4642-99b2-f8d3d4a0cf1a",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "def plot_nz(df, bins=np.arange(0,5,0.2)):\n",
+    "    kwargs=dict( bins=bins,alpha=0.5)\n",
+    "    plt.hist(df.zs.values, color='grey', ls='-' ,**kwargs)\n",
+    "    counts, _, =np.histogram(df.z.values, bins=bins)\n",
+    "    \n",
+    "    plt.plot((bins[:-1]+bins[1:])*0.5,counts, color ='purple')\n",
+    "    \n",
+    "    #plt.legend(fontsize=14)\n",
+    "    plt.xlabel(r'Redshift', fontsize=14)\n",
+    "    plt.ylabel(r'Counts', fontsize=14)\n",
+    "    plt.yscale('log')\n",
+    "    \n",
+    "    plt.show()\n",
+    "    "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 58,
+   "id": "fa8a38f9-d741-489b-aaaa-997f0671a1cc",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjkAAAGzCAYAAADNKAZOAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAABM9UlEQVR4nO3dd3hUVeI+8PdOTZ1JJ4VUgWCA0DtoQFzFXRFdFssKq64FN6sou9jLWnbFhgXzUxGFdbEX4GsBAQUpBgkQCJBQk5AeUmdSJlPv74+YkUiAZDKTO+X9PE+eNVPufTOu5OXcc88RRFEUQURERORlZFIHICIiInIFlhwiIiLySiw5RERE5JVYcoiIiMgrseQQERGRV2LJISIiIq/EkkNEREReiSWHiIiIvJJC6gBSsdlsqKioQHBwMARBkDoOERERdYMoimhqakJsbCxksvOP1fhsyamoqEB8fLzUMYiIiMgBpaWl6N+//3lf47MlJzg4GED7h6TRaCROQ0RERN2h1+sRHx9v/z1+Pj5bcjouUWk0GpYcIiIiD9OdqSaceExEREReiSWHiIiIvBJLDhEREXkllhwiIiLySiw5RERE5JVYcoiIiMgrseQQERGRV2LJISIiIq/EkkNEREReyedKTlZWFtLS0jB27FipoxAREZELCaIoilKHkIJer4dWq4VOp+O2DkRERB6iJ7+/fW4kh4iIiHwDSw4RERF5JZ/dhZx6Z+vWrU47VkZGhtOORURE1IEjOUREROSVOJLjBU5uOglDnQFJ05IQ1C9I6jhERERugSXHw9WfqMfqK1YDv9wj1y+9H5IvS0bKjBQkXpIIVZBK2oBEREQSYcnxcIWbCwERUPgrYDFYUJ1Xjeq8aux6ZRdkChn6T+yPlBkpSL4sGXHj4iBXyqWO3CWryQp9uR66El2nL31p+2NypRw3fnUjNP15uz8REXUPS46HK95aDACY8tAUjLl7DIq3FKNwcyEKNxeisagRJdtLULK9BFuf3ApVkApJGUlIntE+0hOZFglBEFyeURRFmHVmGE8bYTxtRFt1W/s/VxvRdroNexv3ormq2T4adS7ZS7NxxdIrXJ6XiIi8A0uOBxNF0V5ykqYlITAyEEPmDsGQuUMAAA2FDSj8vhBFm4tQ+H0hDHUGHPv6GI59fQwAEBQdZL+0lTIjpVujJDarDW2NbTCUG2BptsDcZIa12QpzkxmWZgssTRb7/5r1Zhhr2suMzWS74LHlajm0Cdr2r3gtNAkaaBO0aGtow6bFm5D7Xi6mPT2Nl+CIiKhbuOKxB6x4fK7btVtOtWDPLXsgU8kw+avJkKnOfbOcaBMxOHQwCje3l55T20/BYrB0ek14ajiSpydD4a9AW2Mb2hravwwNBvv3Rr3R4Z9DFaaCOkoNdT811JFq+PXzgzpKjclXTYY2QYuAyIAuR5ZEm4g3Ut9A/Yl6/P7N32PMgjEOZyAiIs/Wk9/fHMnxYI25jQAAzVDNeQsOAAgyATEjYxAzMgaTF0+Gpc2C0uxSFH1fhMLNhajIqUDd0TrUHa3r1rllfjIoghRQBiuhCFZAEfTLV/AvX4EKKDVKqCJV8IvygzpSfc6MsWNiL5h9bOZYfHf/d9j9xm6Mvmt0n1xmIyIiz8aS48F0+3UAgJARIT1+r8JPgeRpyUielozpz05HW2MbircWo2RHCSAA/qH+8Av1a//fEL/O/xzih+0/bXfyT3N+I24dgR8e+wE1h2tQvKUYydOT+/T8RETkeVhyPJQoimg80AjAsZLzW34hfhg8ezAGzx7c62O5gp/WD8PnD8eeN/dg9xu7WXKIiOiCuOKxh2otboW50QyZWobgwcFSx+kTYzPHAgCOrjuKxlON0oYhIiK3x5EcD9W4vxHAL/NxlJ7dVXuyD1bIyBA05jbiy4e/RMqdKWc9z32wiIiog2f/dnRAVlYW0tLSMHbsWKmj9EpHyXHGpSpPEnddHACg8ptKWI1WidMQEZE787mSk5mZifz8fOTk5EgdxWGiTYTugOOTjj1Z+MRwqPupYdFbcPr701LHISIiN+ZzJccbtJ5qhVlnhsxPhuBU35iP00GQC4i9pv2W84o1FfDRZZ6IiKgbWHI8UMf6ONqhWo+fj+OImKtiIFPJ0HyiGfpDeqnjEBGRm/K935BewFfn43RQapWIuiwKAFC+plziNERE5K54d5WHEW1nrI8zMqRH7+3JXUzuLu66OFStr0LttloYa41QR6iljkRERG6GIzkepqWoBRa9BTI/GYIGBUkdRzJBA4KgGaaBaBVR+VWl1HGIiMgNseR4GPt8nGFayBS+/a8vbnb77eQVX1XAZr7wLudERORbfPu3pAdy9FKVN4q4JAKqCBXMDWbUbK2ROg4REbkZlhwP0ml9nOEh0oZxAzKFDLFXt99OzgnIRET0Wyw5HqSlsAWWJgvk/nKfWx/nXGL+EANBIaCpoAn6I7ydnIiIfsWS40E6bh3XpmshyAVpw7gJVZgKkRmRANoXByQiIurAkuNBOiYd++r6OOcSd237BOTTW06j5XSLxGmIiMhdsOR4CNEqQpfXPh9HO0IrcRr3oknTIHhwMESziL3v7JU6DhERuQmWHA/RXNgMS7MF8gA5ggdyPs5vdYzm7HlzD6xm7k5OREQsOR7Dvj4O5+N0KTIjEsoQJZrKm3B03VGp4xARkRtgyfEQ9lvHOR+nSzKVDDF/iAEA7F62W+I0RETkDlhyPIBoPWO/Kpacc4qdFQtBLuDUtlOozquWOg4REUmMJccDNJ9ohrXFCnmgHEEDfHe/qgtRR6px8XUXAwB+XvazxGmIiEhqLDkegOvjdN+4e8YBAA5+cBCGeoPEaYiISEosOR6go+TwUtWFJUxJQL/0frAYLMh9L1fqOEREJCGWHDdns9igO8hJx90lCIJ9NCcnKwc2K3cnJyLyVSw5bq4ytxLWFisUQQoEXcT5ON0x7KZh8Av1Q2NxI45/c1zqOEREJBGWHDdXvLUYAOfj9IQyQIlRt48CwNvJiYh8mceXnJaWFixevBhXXnklnnvuOanjOF3xlmIAvFTVU2PuHgMIQOHmQtQU1Egdh4iIJOCWJcdgMECn03XrtUVFRViyZAnWr1+P77//3sXJ+pbNYkPJ9hIAQMjIEGnDeJjQ5FCkXp0KoH1uDhER+R63Kjk2mw2rVq3CoEGDkJv7650xRUVFWLBgAZYvX4758+fj1KlT9ueGDh0KuVyOXbt24Y477pAitstU7quEqdkERbACgSmBUsfxOB0TkA/89wCMeqPEaYiIqK+5Vcmpra3FtGnTUFZWZn/MZrNh1qxZmDt3Lu68807MmzcPN9xwQ6f3lZSU4K233sJTTz2Ftra2vo7tMkVbigAA2uFaCDLOx+mp5MuSEXFxBEzNJuxftV/qOERE1MfcquRERUUhMTGx02PfffcdTpw4galTpwIApk+fjry8POTk/HoJIiEhAf/9738xfPhwHDx4sE8zu9Kpre0jViHDQ6QN4qEEQcC4v7eP5ux+YzdEmyhxIiIi6ktuVXK6kp2djZSUFCiVSgCAXC5HSkoKtmzZctZrY2JikJKS0uVxjEYj9Hp9py93ZjVbcWr7LyWH83Eclj4vHapgFeqP1+PkppNSxyEioj7k9iWnuroaGo2m02NarRbl5eUAgNdeew033XQTvv76a1x11VUIDw/v8jjPPfcctFqt/Ss+Pt7l2Xujcm8lzC1m+If5IzCZ83EcpQ5WY8StIwDwdnIiIl+jkDrAhSiVSvsoTgebzQabrX0l24ULF3brOA8//DAWLVpk/16v17t10elYHyfx0kTOx+mlcZnjsPv13Tj+7XHUn6xH2EVhUkciIqI+4PYjOTExMWddWtLpdIiLi+vRcdRqNTQaTacvd9axPk7StCRJc3iD8EHhGHDlAEDk7eRERL7E7UdyMjIy8MILL0AURQiCALPZjOLiYkybNk3qaC5jNVtRsqN9fZykjCQU1BVInMhzbN26tcvH/S7xAzYAOctzoLhcAbm//LzHycjIcH44IiLqU243ktNxGarDxIkTERsbi+3btwMAtm3bhqSkJIwbN86h42dlZSEtLQ1jx47tdVZXqcipgLnVjICIAEQNiZI6jlcIGx8Gv1g/WFusqN5ULXUcIiLqA25VcmpqarBkyRIAwOrVq3HkyBHIZDKsW7cOK1asQFZWFlauXIk1a9ZAEBybp5KZmYn8/PxOt6C7G87HcT5BJiBudvslzvI15RBF3k5OROTtBNFH/7TX6/XQarXQ6XRuNz/nf5f/D4WbCzFz2UyM+/u4c16CoZ6xNFuQ/ads2NpsGPbiMISNOfcEZF6uIiJyTz35/e1WIzkEWE1WlOz8ZT4OJx07lSJIgZiZMQCAsk/LLvBqIiLydCw5bqY8pxwWgwUBkQGITIuUOo7XiZsTB8iAhpwGNJ9sljoOERG5kM+VHHefeGy/dTwjyeF5R3Ru/rH+iJzaXh7LPuNoDhGRN+OcHBfNyXF0Hs2BfxxA475GDFg4wD5RlpxLn69HbmYuBIWA8R+OhzpSfdZrOCeHiMg9cU6Oh7KZbNAfal/4MGREiLRhvJgmTQPNMA1Ei4jyNeVSxyEiIhdhyXEj+iN62Ew2KEOVCEgMkDqOV4uf276lR+VXlbC0WiROQ0RErsCS40Z0+3UAgJDhIZyP42Lhk8Lh398flmYLqr6tkjoOERG5AEuOG2nc3wgACBkZImkOXyDIBPT/U38AQNnnZRCtPjk1jYjIq/lcyXHXu6tsJhv0hzkfpy/1u6IflFoljNVG1PxYI3UcIiJyMp8rOe66rYM+v30+jipMBf94f6nj+AS5Wo7Y2bEA2hcH9NEbDYmIvJbPlRx31XigEQCgHaHlfJw+FHtNLGQqGZqONkGXp5M6DhERORFLjpuwz8cZHiJpDl+jClWh3xX9AHCrByIib8OS4wY6zcfhpOM+139Of0AA6n6qQ2tJq9RxiIjISVhy3IA+Xw/RLEIVroJ/f87H6WsBCQEInxQOgKM5RETexOdKjjveXdWY2wig/a4qzseRRvz17YsDVm2sgqneJHEaIiJyBp8rOe54d5V9Pg5vHZeMZqgGwYODIZpFVKyrkDoOERE5gc+VHHdjNVqhL2ifj6MdoZU4je8SBME+mlO+thzmVrPEiYiIqLdYciSmP/zLfJwIFfzjOB9HShFTI+AX4weL3oL9/90vdRwiIuollhyJnXmpivNxpCXIhfY7rQDsWroLNqtN4kRERNQbLDkS43wc9xI9MxqKYAXqT9Tj6P8dlToOERH1AkuOhKxtVjQVNAFgyXEXcn85Yme1b/WQ/XK2xGmIiKg3WHIkpD+sh2gRoY5Uwy/WT+o49IvYa2MhV8lRurMUpdmlUschIiIH+VzJcad1cjgfxz2pw9UY9udhADiaQ0TkyXyu5LjTOjkdJYe3jrufif+YCAAo+LIA9SfrJU5DRESO8LmS4y6sBiuajvwyH4f7VbmdqCFRGDBzACACu17ZJXUcIiJyAEuORHSHde3zcfqp4RfN+TjuaNI/JwEA9q/cj9Y6btxJRORpWHIkwv2q3F/StCREj4yGudWMPW/tkToOERH1EEuORHQHdAB467g7EwTBPjdn97LdsLRZJE5EREQ9wZIjgU7zcVhy3NqQuUOg6a9BS3UL8j7IkzoOERH1AEuOBGR+Mox+dzRSH0jlfBw3J1fKMf6+8QDabycXbaLEiYiIqLtYciQgCAICEwMRPTNa6ijUDaPvGA21Ro3aglqc2HBC6jhERNRNLDlEF6DWqDHqzlEAgJ9e+kniNERE1F0+V3LcacVj8hzj7x0PmUKG4i3FqNhbIXUcIiLqBp8rOe604jF5Dm28FkOuHwKAWz0QEXkKnys5RI7quJ388KeHoSvRSZyGiIguhCWHqJtiRsYg+bJkiFYRu17jVg9ERO6OJYeoBzq2eti3fB/aGtskTkNEROfDkkPUAxddcRGihkbB1GzC3nf2Sh2HiIjOgyWHqAfO3Orh59d+htVklTgRERGdC0sOUQ8NvXEogqKD0FTehEOfHJI6DhERnYNC6gBE7mjr1q3nfT7yD5FoXtGMDQ9tQHVQNVShqnO+NiMjw7nhiIioWziSQ+SA2Gtioe6nRltFG/IW58HcZJY6EhER/QZLDpEDFEEKpL+UDmWoEi0nW3DwwYOwtFqkjkVERGdgySFyUED/AAx/eTgUGgWaCppw6JFDsLZxIjIRkbvwuZLDvavImQKTA5H+QjrkgXLoDuhw+MnDsJlsUsciIiL4YMnh3lXkbMGpwRj23DDI/GRo2N2AgmcLIFpFqWMREfk8nys5RK6gHabF0GeHQlAKqN1eiyPPH4FoY9EhIpISSw6Rk4SODkXav9IgyAWc3nQax185DlFk0SEikgrXySFyoohJERj8yGAUPFuAyq8rIfNr/3uEIAhOOT7X3CEi6j6O5BA5WdT0KAz65yAAQPnn5Ti16pTEiYiIfBNLDpELxFwVgwH3DAAAnHr/FEo+KpE4ERGR72HJIXKRuOvikHxHMgCgaHkRyteWS5yIiMi3sOQQuVDCTQlIuDkBAHDitROo2lAlcSIiIt/BkkPkYkm3JSHuj3EAgKMvHkXN1hqJExER+QaWHCIXEwQBF2VehOjfRwM2oODZAtRl10kdi4jI67HkEPUBQRAw6P5BiLosCqJVxOEnD6NhX4PUsYiIvBpLDlEfEeQCUh9KRfiUcIhmEYcePQTdIZ3UsYiIvBZLDlEfkilkSHs8DaFjQmFrs+HgQwfRdKxJ6lhERF6JJYeoj8lUMgx5Zgi06VpYW6zIW5yHlqIWqWMREXkdlhwiCcj95Bj6n6EIHhwMi96CvH/mwVBukDoWEZFXYckhkogiUIFhzw9DYEogTPUmHHnuCDf0JCJyIp8rOVlZWUhLS8PYsWOljkIEpUaJYUuGQeYng/6wHrXbaqWORETkNXyu5GRmZiI/Px85OTlSRyECAKgj1Yi/Ph4AULi8EDazTeJERETewedKDpE7ir8+HqowFdoq2lCxrkLqOEREXoElh8gNyP3lSLotCUD7ruXmJrO0gbxca20rvv37t3gh/AUceP+A1HGIyEVYcojcRPSV0QhMDoSlyYKSD0qkjuOVLEYLfnrpJ7w+4HXkZOXAUG/A1ie3QrRxwjeRN2LJIXITglxA8l3JAIDyL8thqOQt5c4iiiIOf3YYWRdnYdPiTTDqjIgeEQ21Ro3G4kYUbSmSOiIRuQBLDpEbCRsXhpDRIRDNIopW8BevM5TvLsfKKSvx+dzP0VjUiODYYFyz8hrcsecODL1xKAAg991ciVMSkSuw5BC5EUEQcNGCiwABqPmhBvp8vdSRPJauRIcv//wlVoxfgdKfSqEMUOLSf12Kvx/7O0bcMgIyuQwj/zoSAFDwZQEMDRw5I/I2LDlEbiZoQBD6XdEPAHDyrZNcILCHjE1GfP/o93gj9Q0c/PAgIAAjbhmBvx/7OzKezIAqUGV/beyYWEQNi4LVaG1/LRF5FZYcIjeUfFsyZGoZ9Af1qNtZJ3Ucj2Cz2LB3+V4sG7AMO/6zA5Y2C5IyknDnnjtxzcproInTnPUeQRDsozm8ZEXkfVhyiNyQOlKN/n/qDwAofLsQNgsXCDyfkxtP4u2Rb+Pru75Gy+kWhA0Mw/Vrr8f8H+YjZlTMed+bfnM65Co5qnKrUJlb2UeJiagvsOQQuan4G+OhDFXCUGZA5Vf85duVmvwafHDVB1h9xWqcPnQafqF+uPK1K/G3Q3/D4GsGQxCECx4jIDwAg2cPBsDRHCJvw5JD5KYUAQok3ZIEACheVQxLs0XaQG6k5XQLvr77a7yZ/iZOrD8BmVKGCfdPwL0n7sX4e8dDrpL36Hgdl6wOfnAQZgMXYiTyFgqpAxDRucX8PgblX5SjtaQVJR+WAH+QOpF0bBYbqvOqcezrY/jppZ9gajIBAAZfOxgznp+B8IHhDh87+bJkaOI10JfqcWTNEQy7aZizYhORhFhyiNxYxwKBhx89jLLPy6BbooM2QSt1rD5hajGh/OdylOwoQcmOEpRll8HUbLI/HzMqBr9b+jskXZrU63PJ5DKMuHUEtj29Dbnv5bLkEHkJlhwiNxc+MRzaEVro9uvww6M/4Nr/XSt1JJdoOd2Ckp3thaZ0Rykq91WeNeFarVEjfnI8ht00DMNuGgZBduE5N9018taR2PbMNhR9X4SGogaEJoc67dhEJA2WHCI317FA4L4F+5C3Og/j7xuP2NGxUsfqFVEU0XCywT5KU7K9BHXHzr5VXtNfg4SpCUiY0v4VOSQSMnn3phJu3bq1x7lCRoWgcW8j1j65Fsm3Jdsfz8jI6PGxiEh6LDlEHiA4NRhRl0fh9KbT2PTPTZj/w/xu3TnkLmwWG6oOVNkLTcmOErRUt5z1uqihUYifEm8vNdoEbZ/+nDFXxaBxbyOqN1Qj6S9JEOSe8xkT0dlYcog8RPJfk1G3rQ7FW4tx7OtjSL06VepIF2QxWrBvxT7s+M8ONFU0dXpOrpIjdmysvdDET4qHf5i/REnbRUyJgCJYAWONEQ17GxA2LkzSPETUOyw5RB7Cr58fJtw3ATuf34nND2zGwJkDIVO45yoQVrMVB94/gG1Pb4OuRAcAUGvVSJicYB+piRsbB4Wfe/0RJFPJ0G9GP5SvKUflt5UsOUQezr3+hCGi85ry8BTsW7EPtUdqsW/FPoxZMEbqSJ3YrDYc+ugQtv5rKxpONgAAgmKCMPXRqRh1+ygo1O7/R070VdEoX1OOup11MDWaoApRXfhNROSW3POvgUTUJT+tHzL+lQEA2PrkVhj1RmkD/UK0iTj82WG8OexNrJm3Bg0nGxAQGYDfvfw73HvyXozLHOcRBQdo3yA1aGAQRIuI05tPSx2HiHqBJYfIw4y+azTCBoah5XQLdr6wU9Isoiji6FdH8faot/H53M9RW1ALv1A/TP/PdCwsXIiJiyZC6a+UNKMjoq+KBgBUfVvFXeCJPBhLDpGHkSvluPyFywEA2S9nQ1+m7/MMoiji5MaTeHfCu/h41seoPlANVbAKlz55KRYWLcTUh6dCFeS5l3n6zegHmUqGlqIWNB1puvAbiMgteXzJqaiowJw5c5CUlIQnnnhC6jhEfSL1mlQkTEmApc2CHx77oU/PXfxjMVZdugqrr1iN8t3lUAYoMfmhyVhYtBAZ/8qAn9avT/O4giJIgYhLIgC0j+YQkWdyy4vkBoMBJpMJWu2Fl6/fvn07PvnkE7S0tGDQoEFYtGgRQkJCXB+SSEKCIODyly7HuxPexYH3D2DCfRMQPSLapecs21WGLY9vQeHmQgCAXC3HmLvHYMpDU7CnYA92H9zd63O406J70VdF4/Tm0zj9w2mYWkxQBXruyBSRr3KrkmOz2fD+++/j8ccfx//+9z/7H3hFRUV4/vnnMWrUKOzYsQPPPPMMEhMTAQDXXXcd5HI5NBoN0tLS4O8v7TobRK7021V8I6dHouaHGnzy10+Q/lJ6jxbO626hqMytxJbHt+D4N8cBADKlDKNuH4Wpj06FJk7T/qKCbp/WY4QMD4FfrB/aKtqQ/3k+RvxlhNSRiKiH3Krk1NbWYtq0aSgrK7M/ZrPZMGvWLLz22muYPn06kpOTccMNNyA7OxsAoFS2T2qsqanBjBkzoFarJclOJIXk25NRu70WjfsaUb+7HuHjHd+J+0wWowVlu8qwe9luFHzR3mAEuYDhfxmOSx+/FCFJIU45jzsTZAKiZ0aj+N1i5L6by5JD5IHcquRERUWd9dh3332HEydOYOrUqQCA6dOnY/bs2cjJycHYsWMBtE+C/Oqrr/Dggw+e89hGoxFG46+32+r1fT9Zk8jZ/GP8EXdtHMo+LUPhm4UIGxPm0FYENosNlfsqUfRDEYp+KELJjhJYDJb2JwVg2E3DcOmTlyJ8oHNKlKeIviIaxSuL7XtrhQ/yrZ+fyNO5VcnpSnZ2NlJSUuwjNnK5HCkpKdiyZYu95KxZswbXX3895HI5Tp06Zb+UdabnnnsOTz31VJ9mJ+oLCTcnoGp9FVpPtaJqfRVi/hBzwfeINhHVedX2UnPqx1NnrbkT2C8QA64YgEkPTELUkLP/AuIL1JFqhI0NQ/3P9chdmYsZz82QOhIR9YDbl5zq6mpoNJpOj2m1WpSXlwMA3nzzTSxZsgTBwcEwmUx44403uiw5Dz/8MBYtWmT/Xq/XIz4+3rXhifqAMliJxPmJOJl1EsUrixF1WRTk/vJOrxFFEYZyAxr3NaIxtxGN+xuxrXFbp9f4hfghKSMJSdOTkDw9GZFpkX2+CagjO4e7WvTMaNT/XI8D/z2A6c9Md9utNIjobG5fcpRKpX0Up4PNZoPNZgMA3H333bj77rsveBy1Ws35OuS1Yq+JRfmacrRVtKH0k1Ik3ZKEttNtv5aa3EYYazqP1CgDlEi8JNFeaqJHREMm5y/w3wqfFI6AyAA0Vzbj+PrjHrExKhG1c2rJMRgMKCgoQGpqKgIDA51yzJiYGOzYsaPTYzqdDnFxcU45PpE3kCllSLkzBfn/ykfpx6U4vfk0DOWGTq8RlAI0aRqEjgpFyMgQ/OGuP0Cukp/jiNRBppQhfV46di3dhdx3c1lyiDyIwyVn0aJFkMvluPbaazFp0iQUFxdj6tSpqKioQHh4OL799luMGdP7zQMzMjLwwgsvQBRFCIIAs9mM4uJiTJs2rdfHJvImEZdEQDNEA/1hfXvBkQHBqcEIGRmC0FGh0AzRQO73a6lhwem+UX8dhV1Ld+HY18fQXNWMoOggh4/lzEty7rSuEJE7crjkfPrpp/j++++Rmtr+t5rbbrsNdXV1+OijjxAfH4+XX34ZH330UY+P23EZqsPEiRMRGxuL7du345JLLsG2bduQlJSEcePGOZQ7KysLWVlZsFqtDr2fyF0JgoCLH78Y1ZuqEXRRELTDtFAEuf0VaY8QmRaJ/hP6o2xXGQ68fwCTH5gsdSQi6gaHL8Dfcccd9oLz0UcfYevWrVi8eDHmzp2LiRMnYuTIkT0+Zk1NDZYsWQIAWL16NY4cOQKZTIZ169ZhxYoVyMrKwsqVK7FmzRqHJ0RmZmYiPz8fOTk5Dr2fyJ359fND4s2JCJ8YzoLjZCP/2v5nWu57udy0k8hDOPynoJ9f+/40NTU1+Oc//4mEhAQ88MAD9ucLCwt7fMzIyEg88sgjeOSRRzo9PmjQILz//vsA2ksKEVFfGzJ3CDYs3IC6o3Uo/akUCZMTpI5ERBfgcMkxmUy49tprsW/fPuh0OmzcuBGBgYEwm8145513sHLlSrz11lvOzEpETuSOt2u7M7VGjSFzh2D/qv3IfTeXJYfIAzh8uerxxx/Hrbfeivvuuw95eXmYNGkSzGYzli1bhpqamrNGY4iIPF3HJavDnx6Gscl4gVcTkdQcHslpamrCrFmzOj2mVCqxaNEi/PTTTxg/fnyvw7kCJx4TkaPiJ8cjPDUcdUfrcPiTwxh1+yipIxHReThcchYuXIj33nuvy+fCwsJw//334/XXX3c4mKtkZmYiMzMTer0eWq1W6jhE5EEEQcDI20Zi84ObkfturuQlx1mXHHkrOnmrHl2uamlpQUlJCUpKStDc3IzS0lL79x1fJ0+exObNm/Hf//7XVZmJiCQzfP5wCHIBZbvKUJNfI3UcIjqPHo3kNDU14Y477sDmzZsBAF988UWXrxNFERMmTOh9OiIiNxMUHYRBfxiEo+uOIve9XPzupd9JHYmIzqFHJSc6OhobNmzAwoULsX37dsyePfus18jlcsTHx+Paa691VkYiIrcy8q8jcXTdURx4/wAu+89lXD2ayE31eE6OIAh4/fXXsWnTJlx++eWuyERE5NYGzhyIoOggNFc149jXx3DxdRdLHYmIuuDwxOMLFZx169bhmmuucfTwLsO7q4iop7qa4Bs6LRTNHzVj8wubUR1W3fehiOiCBNHB9cltNhs+/fRT7N69GzqdrtMy5xaLBZs3b0ZFRYXTgjpbx91VOp0OGo3G6cfnQmtE3q21tBU583MAGTDh4wlQR6qljuQw3l1FnqQnv78dHsm5/fbbsWrVqnM+7+jeUkREniAgPgDadC10eTpUfVeFxJsTpY7kMO6MTt7K4RWPv/jiCzz66KNoaGiAzWbr9KXT6XDXXXc5MycRkduJnhkNAKhaXwXRxk07idyNwyUnJSUFt956a5cL6gUHB9t3Eyci8laRl0ZCHiBHW0UbdHk6qeMQ0W84XHJefPFFfPnll+d8fuPGjY4emojII8j95YiaHgUAqPy2UuI0RPRbDs/J2bhxIzZu3Ii8vDwoFJ0PY7PZsGnTJsyZM6fXAZ2Nd1cRkTNFXxWNyq8rUftjLSz3WqAIcviPVSJyMof/azx69Cjy8vKQl5fX5fPuOvGYe1cRkTMFDw5GQFIAWotbUfZFGZL+kiR1JCL6hcOXq+bMmYNdu3adNenYZrPBYDDgvvvuc2JMIiL3JAgCEue131lV8kEJWktaJU5ERB16VXKGDBnS5XMqlQqPPfaYw6GIiDxJ5LRIhI0Pg2gWcezlY7zTishNOFxy/P39ERgY2OVzu3fvxr59+xwORUTkSQRBwMD7BkLmJ4MuT4fKbzgJmcgdODwnJyUl5ZzPVVdX469//Ssuu+wyRw9PRORR/KL9kPzXZJzMOonCtwsRPikc6nDPXQWZyBs4XHJqa2sxYsQIyOWdd9/V6/UIDw9HUFBQr8MREXmSuGvjcPr702g60oQTr5/AkKe6vqRPRH3D4ZLz6quv4rbbbuvyucWLF2PWrFkOhyIi8kSCXMCgfwzC3rv2onZbLWp31CJiSoTUsYh8Vq8mHp/LDTfcgIULFzp6aJfKyspCWloaxo4dK3UUIvJCQQOCEH9DPADg+GvHYWmxSJyIyHc5XHLOt/PnwYMHcfjwYUcP7VKZmZnIz89HTk6O1FGIyEslzk+Ef5w/TLUmFL1TJHUcIp/l8OWq6dOnd/l4a2sr9u7di6lTpzociojIk8nVcgz8x0DkLcpDxboKRF0WBe0wLj5K1NccLjk//vgjYmNjz9rSQaVS4brrrsOLL77Y63BERJ4qdGQoomdGo2p9FY69fAyjl4+GTOXw4DkROaBXE4/vueceZ2YhIvIqKQtSULerDq2nWlHyUQm3fCDqYw7/tWLBggX2f25ra0NVVRUsFk6wIyLqoNQoMeDvAwC0b/nQcqpF4kREvsXhkqNUKnHw4EFcdtllCA4ORlxcHDQaDebOnYuSkhJnZiQi8liR0yIRNuGXLR9e4pYPRH3J4ctVBQUFmDx5MoxGI4YPH46kpCRYrVYcOHAA48aNw+7du5GQkODMrEREHqdjy4c9t+6B/pAelV9XInZWrNSxiHyCwyM5jzzyCK688kocP34ce/bsweeff441a9bg6NGjeOONN/D00087MycRkcfy6+eHpL8mAQAKlxfCWGOUNhCRj3C45NTX1+Pjjz/ucrRmzpw5UKlUvQrmKlwMkIikEDc7DsEXB8PaYsWJZSekjkPkExwuOSNHjoRMdu6319XVOXpol+JigEQkBUEuYNA/B0GQC6jdXova7bVSRyLyeg6XnLq6OhiNXQ+5fvjhhygq4iqfRERnCkoJQvyNZ2z50Mw7UolcyeGJx1dccQUmTZqEv/3tb0hMTITRaERBQQHWrl2L7OxsfPzxx87MSUTkFRLnJaLmxxoYSg0oXF6IQYsGSR2JyGs5XHJuvvlmHDt2DAsWLIDNZgMAiKIItVqNpUuX4k9/+pPTQhIReQuZSoZBiwbhwP0HUPlVJaJmRCEkPUTqWEReyeGSAwBPP/00brvtNmzcuBG1tbWIjY3FlVdeiejoaGflIyLyOiEjQhD9+2hUfVOF4y8fx+h3uOUDkSt0u+SkpaWhtbUVBoMBY8aMweOPP44JEyYgKSkJd955p/1177zzDmw2G+666y6XBCYi8gYpd6Wg7qc6tJa0ouSDEiTdmiR1JCKv0+2/Ohw5cgSnT5/G22+/jW+++QYTJkzo8nV33HEHfvzxRxw9etRpIYmIvI0yWImB9w4EAJR8WIKWIm75QORsPRofXbp0KWbPnn3B1z300EN44403HM1EROQTIi6NQPikcIgWEcde5pYPRM7W7ZKj1Wpx++23d+u16enp+Pnnnx0ORUTkCwRBwICFAyD3l0N/WI+K/6uQOhKRV+l2yUlOToZC0f15yk1NTQ4FIiLyJX5Rfki+IxkAUPROEbd8IHKibpectra2bh/UbDbj9OnTDgUiIvI1sdfEQjNEA2urFcdfPQ5R5GUrImfodsmRy+U4cOBAt167fv16REZGOhyKiMiXCDIBg/4xCIJCQN1Pdajdxi0fiJyh2yVn5syZuPPOO2EwGM77urq6OixevBgzZszodThX4AadROSOApMDf93y4dXjqPiqAlaDVeJURJ5NELs5LlpdXY3U1FT0798fL7zwAmbOnAlBEOzPm81mrF27Fg8++CAqKytx8OBBDBgwwGXBe0uv10Or1UKn00Gj0Tj9+Fu3bnX6MYnIu9lMNuy7ex9aCttvJ1cEKRD9+2jEzY6DX7SfxOm6JyMjQ+oI5OV68vu72zOJ+/Xrh08++QSzZ8/G1VdfjcDAQKSmpsLf3x86nQ7Hjh2DyWQCAKxYscKtCw4RkTuSqWQY8foIVK2vQvmacrRVtKHskzKUfVaG8InhiLsuDiEjQzr9BZOIzq3bIzkdcnNz8be//a3LW8QTExPx2muvYdasWU4L6CocySEidyZaRdTvrkf5l+Vo2NNgfzwgKQBx18Wh34x+kPvLJUzoehwVoq705Pd3j0tOh0OHDiE7Oxu1tbUIDg7G8OHDMWnSJMjlnvEfHUsOEXmKllMtqFhTgarvqmBra98Q2RMvZfUUSw51pU9KjqdjySEiT2NptqBqw6+XsgAAMnjtpSyWHOqKS+bkEBGRtBRBCvSf0x9x18Z1upRVt7MOdTvr2i9lXRuHfpc7filLtIowNZpgqm//MjeYYao3wdJsQeSlkQhODXbyT0XkOiw5REQeRpALCJ8YjvCJ4Z0uZbUWt+L4K8dR9E5Rp0tZok2EWW+2FxZTvQmmhrOLjKneBLPODJxjfL/8i3IM/c9QhI4O7dsfmMhBvFzFy1VE5AW6vJQlAKowFUwNJsDWg4PJAFWICspQJVRhKqjCVDBUGKA/qIdMLcPQ54YidKTriw4vV1FXeLmKiMjHnOtSlqnO9OtrNAp7aVGFquz/fGaZUYWqoNQqIcg7z+2xmWw4/MRh1P9cj0OPHMKw54YhZERIH/+URD3DkkNE5EXOvJRlqDTAorfYi4xM0e1F7s8iU8kw5OkhOPT4ITTsbsDBhw9i2PPDEJIe4rzwRE7m+P/jiYjIrfnH+CM4NRjqSHWvCk4HmUqGoc8MReiYUNjabDj44EHoDuqckJTINVhyiIio22QqGYY8OwQho0N+LTqHWXTIPXHiMSceExH1mLXNikOPHEJjbiPkAXKkv5gOTZrz/yx1Fk5i9h49+f3NkRwiIuoxuZ8cQ/89FNoRWlhbrch7IA/6Ar3UsYg6YckhIiKHyP3lGPafYdCma2FtsSJvcR6ajjZJHYvIjiWHiIgcJveXY9iSYdAM07QXnX/moekYiw65B58rOVlZWUhLS8PYsWOljkJE5BXsRWeIBpZmS3vROc6iQ9LzuZKTmZmJ/Px85OTkSB2FiMhrKAIUGPb8MASnBcPS1F50mk80Sx2LfJzPlRwiInINRaAC6c+nI3hwMCx6Cw784wCaC1l0SDosOURE5DSKIAXSX0xHcGp70cn7Rx5ailqkjkU+iiWHiIicShGkwLAXhyFoYBDMjWYc+McBtBSz6FDfY8khIiKnUwYrkf5SenvRaTDjwKIDaC1plToW+RiWHCIicgmlRon0F9MReFFge9G5n0WH+ha3deC2DkRELmXWtY/ktBS2QBWuwvBXhyOgf4DUsRzC7SGkx20diIjIbSi1SqS/nI7A5ECY6kztIzqlHNEh12PJISIil1OFqJD+cjoCEgNgqjUh9++5aNzfKHUs8nIsOURE1CdUoSoMf2U4gi/+5fbyxXmo2lAldSzyYiw5RETUZzqKTmRGJESLiKPPH0XhO4UQbT45PZRcjCWHiIj6lFwtx8WPX4yEeQkAgNIPS5H/VD6sbVaJk5G3YckhIqI+J8gEJN+WjMEPD4agFFC7rRYH7jsAY51R6mjkRVhyiIhIMv1+1w/DXx4OhUaBpqNNyP1bLjf2JKdRSB2AiIh8m3aYFqP+3ygceuQQWktasf/e/bj48YsRPjFc6mgu5cz11Lh+T9c4kkNERJLzj/PHiDdGIGRUCKwGKw49dghln5fBR9erJSdhySEiIregDFZi2PPDEPOHGMAGnMw6iROvnoBoZdEhx/ByFRERuQ2ZQoaBiwbCP94fhW8VouL/KmCoMCDtyTQogvgr61x46atrHMkhIiK3IggC4ufGY8jTQyDzk6FhTwNy/54LQ6VB6mjkYVhyiIjILUVMicCI10dAFaFC66lW5P4tF7pDOqljkQdhySEiIrcVPDAYo/7fKAQNDIK5sX0389Pfn5Y6FnkIryk5hw4dkjoCERG5gDpSjRGvjUD4lHCIZhEFzxag+L/FvPOKLsgrSs6ePXswYcIEqWMQEZGLyP3lGPLUEMTfEA8AOLXqFI78+whsJpvEycidecVU9TFjxiAiIkLqGERE5EKCTEDKXSnw7++P468cx+nvT8NQbkC/K/pBc7EGgRcFQqbwir+7k5O4ZckxGAwwmUzQarVSRyEiIjcT8/sY+MX4If/JfDQdaULTkSYAgEwlQ9CgIGgu1iD44mBo0jRQR6khCILTzu3MW7XdlTfdju5WJcdms+H999/H448/jv/973/2D6eoqAjPP/88Ro0ahR07duCZZ55BYmKitGGJiEgyoaNCMWr5KFRvqIa+QI+mgiZYmi3QH9JDf0hvf50qTNVeeH4pPsGDg6EIcKtffeRCbvVvura2FtOmTUNZWZn9MZvNhlmzZuG1117D9OnTkZycjBtuuAHZ2dkSJiUiIqn5x/gj6dYkAIBoE2EoM9gLjz5fj5bCFpjqTajbWYe6nXXtbxKAwKTATsUnMCkQgtx5oz3kPtyq5ERFRZ312HfffYcTJ05g6tSpAIDp06dj9uzZyMnJwdixY7t9bKPRCKPRaP9er9ef59VERORJBJmAgIQABCQEIPqKaACA1WhF87Hm9uKT3wT9ET2M1Ua0FLWgpagFVd9WAWif1ByUGoSIyRHoP6e/lD8GOZlblZyuZGdnIyUlBUqlEgAgl8uRkpKCLVu22EvOvn37UFNTg02bNuHyyy/v8jjPPfccnnrqqT7LTURE0pKr5dAO00I77Nf5ncY6Y/tIzy8jPk1HmmA1WKHbr4Nuvw4ho0IQlBIkYWpyJrcvOdXV1dBoNJ0e02q1KC8vt38/atQotLS0nPc4Dz/8MBYtWmT/Xq/XIz4+3rlhiYjIranD1VBPUSNiSvsduaJVRMupFhxfehz6w3o05DSw5HgRt7/XTqlU2kdxOthsNthsPVsbQa1WQ6PRdPoiIiLfJsgFBKUEITIjEgDQsKdB4kTkTG5fcmJiYs6aP6PT6RAXFydRIiIi8jahY0MBAI0HGmE1WiVOQ87i9iUnIyMDRUVF9uW7zWYziouLMW3aNImTERGRtwhICIA6Sg3RLEKXx01AvYXblZzfXoaaOHEiYmNjsX37dgDAtm3bkJSUhHHjxjl0/KysLKSlpfXoziwiIvJugiAgdEz7aE5DDi9ZeQu3Kjk1NTVYsmQJAGD16tU4cuQIZDIZ1q1bhxUrViArKwsrV67EmjVrHF7BMjMzE/n5+cjJyXFmdCIi8nD2ksN5OV5DEH10G1e9Xg+tVgudTueSSci+sPQ3EZE3MevM+OnanwARmPDZBKgj1FJH8niu2NahJ7+/3Wokh4iISCpKrRLBqcEAOJrjLVhyiIiIftFxlxXn5XgHnys5nHhMRETnYp+Xs7cBos0nZ3N4FZ8rOZx4TERE56JJ00AeIIdZZ0bziWap41Av+VzJISIiOheZQoaQkSEAeMnKG7DkEBERnaHjklX9nnqJk1BvseQQERGdIWxsGABAf0gPq4FbPHgylhwiIqIz+MX6wS/GD6JFROP+RqnjUC/4XMnh3VVERHQ+giDwVnIv4XMlh3dXERHRhYSNab9kxXk5ns3nSg4REdGFhIwMAWSAodSAtqo2qeOQg1hyiIiIfkMRpIDm4vZ9kRr28pKVp2LJISIi6kLHvJz6HF6y8lQsOURERF3ouJW8cW8jRCu3ePBEPldyeHcVERF1R3BqMBRBCliaLWg62iR1HHKAz5Uc3l1FRETdIcgFhIwKAQA07OG8HE/kcyWHiIiouzouWXFejmdiySEiIjqHjn2s9Pl6WJotEqehnmLJISIiOge/aD/4x/sDNnCLBw/EkkNERHQeHaM53OLB87DkEBERnYd9Xo4EWzzYLLY+P6c3YckhIiI6j5ARIRAUAtoq2mAoN/TZeRsPNGLHVTtQ9F5Rn53T2/hcyeE6OURE1BNyfzk0Q37Z4qGPbiUXRRFF7xRBNIuoWl8FUeRihI7wuZLDdXKIiKinOubl9NUlK90BHfSH9QAAU60JrcWtfXJeb+NzJYeIiKin7Fs87Gvsk3kyp/53qtP3XKfHMSw5REREFxA0MAgKjQLWViuaCly7xYM+X4/GfY0Q5ALi/hgHgCsuO4olh4iI6AIEmfDrJSsXj6qcWt0+ihN1eRRifh8DoP3yldVodel5vRFLDhERUTeEjWm/ZOXKUZXmE82oz64HBCDhpgQEJAVAFaGCzWSD7qDOZef1Viw5RERE3dAxktN0tAlmvdkl5yj5oAQAEJkRiYD4AAiC0Cflylux5BAREXWDOlKNgKSA9i0e9jU6/fitJa2o+bEGAJDw5wT746FjueKyo1hyiIiIusmVt5KXfFgCiED4pHAEXRT06zlHhwIC0FLYAmOd0enn9WY+V3K4GCARETmq41byhpwGpy7Q11bVhupN1QCAhJsTOj2n1CoRNLC99PCSVc/4XMnhYoBEROQobboWglKA8bQRhlLnbfFQ8lEJYGsfKdJcrDnreXu5YsnpEZ8rOURERI6S+8mhTdcCcN6t5MZaI6rWVwHoPBfnTPad0Pc2QLRxi4fuYskhIiLqAWff7VT6SSlEswjNUA20w7VdvkYzRAOZnwzmBjNaCluccl5fwJJDRETUAx2jKo37G2Ez9W6LB7POjMqvKwEAifMSIQhCl6+TKWUIGRkCgFs89ARLDhERUQ8EXhQIZagStjYbdId7t0Bf2edlsLXZEDQoyH6r+LnYR5B4K3m3seQQERH1gLMW6LM0W1C+phwAkHjzuUdxOnSUIN0hHawGbvHQHSw5REREPeSMBfrK15bD2mJFQFIAwieHX/D1/v39oe6nhmgW0ZjX6PB5fQlLDhERUQ+Fjm4vOc3Hm2FqMPX4/VaDFWWflQFo36NKkJ1/FAf4zQgSL1l1C0sOERFRD6nCVAga8MsCfXt7XjgqvqqARW+BX6wfoqZHdft99hEkrpfTLSw5REREDrCvXdPDwmEz2VD26RmjOPILj+J0CBkVAsiA1lOtaDvd1qPz+iKWHCIiIgecWXJ6ssVD1YYqmOpMUEeq0e93/Xp0TmWwEsGDg+3npfNjySEiInKAdpgWMrUMpjoTWoq6t0CfzWJr34gTQPwN8ZApe/5r2NmLEXoznys53KCTiIicQaaSIWRECIDuF47Tm0/DWG2EMlSJ6N9HO3TeTls8WLnFw/n4XMnhBp1EROQs9sLRjbudRKtoH8Xp/6f+kKvlDp1Tk6aBPFAOi96CpuNNDh3DV/hcySEiInKWjpKjy9PBajz/An0122pgKDVAEaxA7DWxDp9TkAsIHcW7rLqDJYeIiMhBAYkBUEWoYDPZoMs79xYPoiiiZHX7KE7cH+OgCFD06rw9GUHyZSw5REREDhIEAWFjLzwRuC67Di2FLZD7yxF3bVyvz9tRcvSH9bC0WHp9PG/FkkNERNQLF1ov58xRnNhrYqHUKHt9Tv9Yf/jH+UO0imjc39jr43krlhwiIqJeCB0dCghAS2ELjHXGs55v3NeIpoImyFQy9P9Tf+ed18HFCH0JSw4REVEvKLVKBA/6ZYG+LubInFp9CgAQ8/sYqMJUTjsvS86FseQQERH10rn2lNId0kG3XwdBIaD/Dc4bxQGAkJEhgAwwlBlgqDQ49djegiWHiIiolzot0Gf7dYG+jrk4/a7oB78oP6eeUxGogGaIpv28vMuqSyw5REREvaRJ00DuL4e50Yzmk80AgKZjTaj/uR6QAQk3JrjkvN25s8uXseQQERH1kkwpa798hF9HVUo+aB/FiZoeBf84f5ec1z6CtI9bPHSFJYeIiMgJzpwI3FLcgtpttQCAhJtcM4oDAMGDgqEIVsDaYoW+QO+y83gqlhwiIiIn6Jh8rDuoQ/F7xQCAiKkRCEwOdNk5ucXD+bHkEBEROYF/nD/8ov0gWkTUbv9lFOfPrhvF6XCuO7uIJYeIiMgpBEGwX7ICgNBxoQhODXb5ee1bPBToYWnmFg9nYskhIiJyko5RFQBIvDmxT87p188PAQkBgK19AjL9yudKTlZWFtLS0jB27FipoxARkZcJGxsGzVANon8fDe0wbZ+dl7uSd83nSk5mZiby8/ORk5MjdRQiIvIycn85Ri4bidR/pvbpeTtGkOpz6iGKvJW8g8+VHCIiIm8TMjwEgkKAsdoIQzm3eOjAkkNEROTh5P5y++UxXrL6FUsOERGRF+Cu5GdjySEiIvICHSWnMbcRNrNN4jTugSWHiIjICwQNCIIyRAmrwQp9Prd4AFhyiIiIvIIgExA6mpeszsSSQ0RE5CW4Xk5nLDlEREReoqPkNB1rgllnljiN9FhyiIiIvIQ6Qo3AlEBABBr2cjSHJYeIiMiL8FbyX7HkEBEReZEzS46vb/HAkkNERORFtOlayFQyGGuMaD3VKnUcSbHkEBEReRG5Wg5t+i9bPPj4JSuWHCIiIi/TsSs5Sw4RERF5FfsWD/sbYTP57hYPLDlEREReJjA5EKpwFWxGG3QHdVLHkQxLDhERkZcRBIG3koMlh4iIyCux5LDkEBEReaWOzTqbTzTDVG+SOI00WHKIiIi8kCpUhaCBQQB8d4sHlhwiIiIv5eu7krPkEBEReamO9XLq99T75BYPHl9yzGYznnjiCaxduxYvvfSS1HGIiIjchnaIFjI/GcwNZrQUtkgdp8+5ZckxGAzQ6bp3X/+KFSswYMAAzJ49G01NTdi5c6eL0xEREXkGmUqGkOEhAHzzLiu3Kjk2mw2rVq3CoEGDkJuba3+8qKgICxYswPLlyzF//nycOnXK/tyuXbswfPhwAMCIESOwYcOGPs9NRETkrjrm5dTn1EucpO+5Vcmpra3FtGnTUFZWZn/MZrNh1qxZmDt3Lu68807MmzcPN9xwg/35qqoqBAW1zx4PDg7G6dOn+zw3ERGRuwobFwYA0OXpYG2zSpymb7lVyYmKikJiYmKnx7777jucOHECU6dOBQBMnz4deXl5yMnJAQCEh4ejubkZANDc3IyIiIguj200GqHX6zt9EREReTv/eH+oo9QQzSJ0eb61xYNblZyuZGdnIyUlBUqlEgAgl8uRkpKCLVu2AACmTZuGgwcPAgDy8vJw2WWXdXmc5557Dlqt1v4VHx/fNz8AERGRhARBgHa4FgDQdKxJ4jR9y+1LTnV1NTQaTafHtFotysvLAQC33norCgoK8Omnn0Iul2P69OldHufhhx+GTqezf5WWlro8OxERkTuQ+8kBAKLVt24jV0gd4EKUSqV9FKeDzWaDzda+dbxCocC///3vCx5HrVZDrVa7JCMRERG5H7cfyYmJiTlr/oxOp0NcXJxEiYiIiMgTuH3JycjIQFFRkX2lRrPZjOLiYkybNk3iZEREROTO3K7kdFyG6jBx4kTExsZi+/btAIBt27YhKSkJ48aNc+j4WVlZSEtLw9ixY3udlYiIiNyXW83JqampwTvvvAMAWL16NaKjozF48GCsW7cOzz77LA4ePIjs7GysWbMGgiA4dI7MzExkZmZCr9dDq9U6Mz4RERG5EUH0xR27AHvJ0el0Z9295Qxbt251+jGJiIgccWzpMVR+VYnEWxKR9JekPjtvRkaG04/Zk9/fbne5ioiIiMgZWHKIiIjIK/lcyeHEYyIiIt/gcyUnMzMT+fn59r2viIiIyDv5XMkhIiIi38CSQ0RERF6JJYeIiIi8EksOEREReSW3WvG4L2RlZSErKwsWiwUAztr801laWlpcclwiIqKeMpgNaEMbDCZDn/5+csXv2I5jdmctY59d8bisrAzx8fFSxyAiIiIHlJaWon///ud9jc+WHJvNhoqKCgQHBzu8D9a56PV6xMfHo7S01CVbRlDX+LlLg5+7NPi59z1+5tL47ecuiiKampoQGxsLmez8s2587nJVB5lMdsEG2FsajYb/IUiAn7s0+LlLg5973+NnLo0zP/fubrDNicdERETklVhyiIiIyCux5LiAWq3Gk08+CbVaLXUUn8LPXRr83KXBz73v8TOXRm8+d5+deExERETejSM5RERE5JVYcoiIiMgrseQQERGRV2LJIa9iMBig0+mkjkFERC5QUVHRo9ez5DhZUVERFixYgOXLl2P+/Pk4deqU1JF8gs1mw6pVqzBo0CDk5uZKHccn/N///R9SU1Oh0Wjwxz/+EfX19VJH8hm5ubmYMmUKwsLCMGPGDNTW1kodyWe0trYiLS0NxcXFUkfxCaIoYtCgQRAEAYIg4Oabb+7R+1lynMhms2HWrFmYO3cu7rzzTsybNw833HCD1LF8Qm1tLaZNm4aysjKpo/iEwsJCfPPNN/jyyy+xatUqbN26FQ8++KDUsXxCW1sbvvjiC2zcuBGlpaVobW3F0qVLpY7lM5YtW4aCggKpY/iM9evX495770VOTg5ycnLw2Wef9ej9Prutgyt89913OHHiBKZOnQoAmD59OmbPno2cnByMHTtW4nTeLSoqSuoIPmXHjh1YtmwZVCoVhgwZgry8vB7/4UOO0el0eOKJJ6BSqQAAU6dOveD+PeQc69atw7Rp06SO4VPeeOMNXH311YiKikJCQkKP38//MpwoOzsbKSkpUCqVAAC5XI6UlBRs2bJF4mREzjV//nz7L1kA6Nevn0N/AFHP9evXz/7Zm0wmVFVV4f7775c4lfcrKSlBZWUlxo0bJ3UUn9HU1ASj0YjHHnsMycnJuOeee9DTpf1Ycpyourr6rE3btFotysvLJUpE1Df27duHu+66S+oYPuXbb7/FhAkTsGXLFhw+fFjqOF7NarXinXfewZ133il1FJ8SHByM77//HlVVVXj11Vfx5ptv4vXXX+/RMVhynEipVNpHcTrYbDbYbDaJEhG5XmVlJSwWC2bPni11FJ9yxRVX4PPPP8ekSZN6PBmTeiYrKwt33XUXLwtKRKlU4p577sFDDz2EDz/8sEfv5b8xJ4qJiYFer+/0mE6nQ1xcnESJiFzLarXi1VdfxbJly6SO4nM6Loe/9957qKmpQU1NjdSRvNayZcswYMAA+Pn5wc/PDwCQmpqKxYsXS5zMt1xzzTU9XiKEE4+dKCMjAy+88AJEUYQgCDCbzSguLuZENfJar7zyChYtWoSgoCAA7XNEzpyrQ64XEBCAiIgIhIaGSh3Fax0/frzT94Ig4OjRo0hKSpImkI+yWq1ITU3t0Xs4kuNEEydORGxsLLZv3w4A2LZtG5KSkjhRrY/wsmDfevXVVzFo0CA0NDTgyJEj+Pbbb7FhwwapY3m9uro6fPXVV/YJmD/++CPmzZsHhYJ/ZyXvsn37dqxevdr+//Xly5f3ePSM/1U4kUwmw7p16/Dss8/i4MGDyM7Oxpo1ayAIgtTRvF5NTQ3eeecdAMDq1asRHR2NwYMHS5zKe3366adYtGhRpzsdAgICUFVVJWEq31BUVITbb78dqampmDNnDoKCgvDvf/9b6lhETldaWor77rsPH330ESZMmID58+djypQpPTqGIPb0fiwiIiIiD8DLVUREROSVWHKIiIjIK7HkEBERkVdiySEiIiKvxJJDREREXoklh4iIiLwSSw4ReTSz2YwffvgBy5cv79H7DAaDixIRkbtgySEil1i/fj3mzp0LQRAgCALS09MxefJkDBw4EKNHj8aTTz6JxsbGXp1Dp9Ph6aefxowZM7q1cZ/RaMSLL76IqVOnXnBTy7a2NgwaNAh//vOfz3quvr4e999/PxYvXozIyEhcc801aGtrc/jnICLXYMkhIpeYOXOmfRVqANizZw927tyJo0eP4t5778Wzzz6L0aNH96roaLVaPPPMM93eBFetVuOOO+7AwYMHYbVaz/taURTP+Zrbb78dcXFxePHFF/Hpp5/CZDLBZDKhpaUFFRUVPf45iMg1WHKIyGW0Wq39nzs27pTJZPjLX/6C6667DoWFhXj33Xd7fR65XN7t14aEhCAkJOSCr/P398fJkyfxwQcfdHq8vLwca9euRUREBABg2rRpWL9+PTQaDRYuXIhjx471KDsRuQ5LDhFJ4qKLLgIAFBcXSxukh06dOoWudsN5+eWXnVLYiMh5WHKISBI5OTkAgBEjRnR6/PXXX8fs2bMxefJkXHzxxfj00087Pd/a2or7778f8+bNwwMPPIDHH38cJpOp02saGxtxzz334JVXXsG9994Lf39/rF279qwMu3btwvjx4xEQEIBx48bhyJEj9uesViu2bNmCp556yv7Yww8/jKeffhoAsGLFCtxyyy1YsmQJ9u3bZz/+kiVLcMstt+DkyZOOfjRE5CwiEZELARDP/KPm+PHj4oIFC0QA4pw5c0SLxWJ/7tFHHxXnzZsn2mw2URRFMTMzU5TJZOKPP/4oiqIo2mw28Xe/+534xBNP2N+TnZ0tAhAvvfRS+2P33XefuHTpUvv3b7/9trhmzRr794mJieLAgQPF++67TywuLhZ37twpBgYGildccYUoiqJYU1MjPvvss6Kfn5+YmJjY6efZsmWLCEBcuXJlp8dXrlwpAhC3bNniyMdERC7AkRwi6hNXX3014uPjMXDgQFgsFvz000/47LPP7PNpysrK8Pzzz+Phhx+GIAgAgJtvvhk2mw1vv/02AOCDDz7Azp078cADD9iPO2HCBMTGxnY618mTJ7Fp0yZYLBYAwLx58xAfH9/pNSEhIXjllVeQmJiISZMmYfr06di9ezcAICIiAo8++ihGjRrlmg+DiPqEQuoAROQbvvrqK7z22mu47777kJ+fj7Fjx3Z6fuPGjbBYLFiyZIm95FgsFgwfPhxKpRIA8PHHH+Oiiy5CYGBgp/d2PN/hxhtvxE033YRJkybhjTfewLhx4zB69OhOrwkICOj0fVhYGBoaGs57XCLyLCw5RNRnFi5ciK1bt2Lt2rV44IEHsHTpUvtzVVVVAIDly5dDrVZ3+f4TJ04gODj4gue58cYb0djYiH/84x+YMGEC7rrrLrz00ktnlSMi8m68XEVEfeq9995DYmIiXnnlFXzxxRf2x6OjowEABQUFZ73HaDTCaDRCq9We8+6mM5lMJtx99904fPgw5s6di7feeqvLRf2IyLux5BCRy3RVRkJDQ/HJJ59AqVTitttuw/HjxwEAY8aMAQA89thjMJvN9tdbLBY89NBDsNlsGDVqFGpqavDdd9+ddZ4zz/XSSy8BAJKTk/Hxxx9jwYIF2Lx5s9N/vjN1zC06MzsRSYslh4hcpra21v7PZ853GT9+PJ577jno9XrMmTMHLS0tSE9Pxx//+Ed88803mDx5Ml5++WUsW7YMl1xyCaZMmQJ/f3889NBD8Pf3x4IFC7Bt2zbU19fj7bffRl1dHfLz8/Hxxx+jra0NP/zwAw4cOGA/X0JCAiZNmmT/3mg0nrV3VWtrKwB02p7BYDCctV1DS0sLAKC5ubnT4x0Tmzdu3IhTp05hw4YNDn1mRORE0t7cRUTeav369eLMmTPtt5DPnj27023cNptNvPrqq0UA4ujRo8X3339fbGpqEhcsWCBGRESIQUFB4qRJk8T169d3Ou727dvF8ePHiyqVSkxOThZXr14tXnXVVeIDDzwg7t+/XxRFUbz00kvF8PBw8aGHHhKfeuop8eabbxYrKyvF1tZW8d///rcIQFQqleLSpUtFo9EorlixQgwNDRUBiPfcc49YXFwsvv7666JKpRIBiP/617/EyspK8auvvhJnzJghAhCHDBkirlq1SmxsbBRFURQtFos4c+ZMMSAgQLz++utFvV7fZ581EXVNEMULXNwmIiIi8kC8XEVEREReiSWHiIiIvBJLDhEREXkllhwiIiLySiw5RERE5JVYcoiIiMgrseQQERGRV2LJISIiIq/EkkNEREReiSWHiIiIvBJLDhEREXkllhwiIiLySiw5RERE5JX+PwI4nHpXCItKAAAAAElFTkSuQmCC",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plot_nz(df_test)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 68,
+   "id": "b8f71544-5a64-4d52-8f50-af5f7fa9929e",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "def plot_photoz(df, nbins,xvariable,metric, type_bin='bin'):\n",
+    "    bin_edges = stats.mstats.mquantiles(df[xvariable].values, np.linspace(0.1,1,nbins))\n",
+    "    ydata,xdata = [],[]\n",
+    "    \n",
+    "    \n",
+    "    for k in range(len(bin_edges)-1):\n",
+    "        edge_min = bin_edges[k]\n",
+    "        edge_max = bin_edges[k+1]\n",
+    "\n",
+    "        mean_mag =  (edge_max + edge_min) / 2\n",
+    "        \n",
+    "        if type_bin=='bin':\n",
+    "            df_plot = df[(df[xvariable] > edge_min) & (df[xvariable] < edge_max)]\n",
+    "        elif type_bin=='cum':\n",
+    "            df_plot = df[(df[xvariable] < edge_max)]\n",
+    "        else:\n",
+    "            raise ValueError(\"Only type_bin=='bin' for binned and 'cum' for cumulative are supported\")\n",
+    "\n",
+    "\n",
+    "        xdata.append(mean_mag)\n",
+    "        if metric=='sig68':\n",
+    "            ydata.append(sigma68(df_plot.zwerr))\n",
+    "            ylab=r'$\\sigma_{\\rm NMAD} [\\Delta z]$'\n",
+    "        elif metric=='bias':\n",
+    "            ydata.append(np.median(df_plot.zwerr))\n",
+    "            ylab=r'Median $[\\Delta z]$'\n",
+    "        elif metric=='nmad':\n",
+    "            ydata.append(nmad(df_plot.zwerr))\n",
+    "            ylab=r'$\\sigma_{\\rm NMAD} [\\Delta z]$'\n",
+    "        elif metric=='outliers':\n",
+    "            ydata.append(len(df_plot[np.abs(df_plot.zwerr)>0.15])/len(df_plot) *100)\n",
+    "            ylab=r'$\\eta$ [%]'\n",
+    "            \n",
+    "    if xvariable=='VISmag':\n",
+    "        xlab='VIS'\n",
+    "    elif xvariable=='zs':\n",
+    "        xlab=r'$z_{\\rm spec}$'\n",
+    "    elif xvariable=='z':\n",
+    "        xlab=r'$z$'\n",
+    "\n",
+    "    plt.plot(xdata,ydata, ls = '-', marker = '.', color = 'navy',lw = 1, label = '')\n",
+    "    plt.ylabel(f'{ylab}', fontsize = 18)\n",
+    "    plt.xlabel(f'{xlab}', fontsize = 16)\n",
+    "\n",
+    "    plt.xticks(fontsize = 14)\n",
+    "    plt.yticks(fontsize = 14)\n",
+    "\n",
+    "    plt.grid(False)\n",
+    "    \n",
+    "    plt.show()\n",
+    "   "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 71,
+   "id": "be87adb7-eb06-433a-8b2c-2cb0c1bb3ae3",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plot_photoz(df_test, 8,'z','bias', type_bin='bin')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 109,
+   "id": "d67d2991-0c55-4ec8-9c5c-8c5cc20364b1",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0.04460718134171633"
+      ]
+     },
+     "execution_count": 109,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "nmad(df_test[df_test.VISmag<25].zwerr)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 110,
+   "id": "a7c59672-ff90-473b-8539-e28d9eaec4b7",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "df_test = df_test[df_test.VISmag<25]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 111,
+   "id": "81e8b270-0da6-41af-b43d-1912fa98ccaa",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0.13433240659117107"
+      ]
+     },
+     "execution_count": 111,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "len(df_test[np.abs(df_test.zwerr)>0.15])/len(df_test)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 112,
+   "id": "5c58a724-ecd8-48ed-89da-0e7eb9a5ca99",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "torch.save(insight.model.state_dict(),'/data/astro/scratch/lcabayol/Euclid/NNphotozs/models/insight_v0.pt')\n",
+    "           \n",
+    "                  "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 113,
+   "id": "c966fd40-3d3f-4df5-a988-c55a1ab2e204",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_test.to_csv('/data/astro/scratch/lcabayol/Euclid/NNphotozs/results/df0.csv', sep=',')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 114,
+   "id": "f8487c14-3b26-4742-a5c0-e6a820400cc9",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/tmp/ipykernel_678/2146862925.py:13: FutureWarning: the 'line_terminator'' keyword is deprecated, use 'lineterminator' instead.\n",
+      "  df_test.to_csv(f, header=True, index=False, line_terminator='\\n')\n"
+     ]
+    }
+   ],
+   "source": [
+    "# Create a list of additional header lines\n",
+    "header_lines = [\n",
+    "    \"# Training spect-zs with a strict quality cut\",\n",
+    "    \"#10 MDN components\",\n",
+    "    \"# For 300 epochs with lr0=1e-3 + 100 epochs with lr=1e-4\",\n",
+    "    \"# Date: 2023-07-26\",\n",
+    "]\n",
+    "\n",
+    "# Write DataFrame to a CSV file with custom header lines\n",
+    "with open('/data/astro/scratch/lcabayol/Euclid/NNphotozs/results/df0.csv', 'w') as f:\n",
+    "    for line in header_lines:\n",
+    "        f.write(line + '\\n')\n",
+    "    df_test.to_csv(f, header=True, index=False, line_terminator='\\n')\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "e20b30c3-00a8-4dd0-969d-0426121b99f5",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "f9c0ad7d-7796-41b5-8b91-7c8dec41f17b",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "60b50881-860c-48c1-80c3-f8a9ab51226f",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c174cd6f-340e-4986-a04d-6318ce11e067",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "DESIenv6",
+   "language": "python",
+   "name": "desienv6"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.10.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

notebooks/match_catalogues.ipynb

@@ -0,0 +1,1126 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "0bd364f7-e6cf-4bd5-906e-2a096ad22ec1",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "from astropy.io import fits\n",
+    "from astropy.table import Table\n",
+    "import numpy as np\n",
+    "\n",
+    "import os"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "id": "cd8362c5-87b5-4b69-bb4b-608d605c2028",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "path = '/data/astro/scratch/lcabayol/Euclid/NNphotozs/Euclid_EXT_MER_PHZ_DC2_v1.5'\n",
+    "filename_calib='euclid_cosmos_DC2_S1_v2.1_calib_clean.fits'\n",
+    "filename_valid='euclid_cosmos_DC2_S1_v2.1_valid.fits'"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "id": "db81717b-09c9-473e-b7c3-488ad3f63b90",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "hdu_list = fits.open(os.path.join(path,filename_calib))\n",
+    "cat_calib = Table(hdu_list[1].data).to_pandas()\n",
+    "\n",
+    "hdu_list = fits.open(os.path.join(path,filename_valid))\n",
+    "cat_valid = Table(hdu_list[1].data).to_pandas()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "id": "066de14f-db52-4419-93e3-4645901bb216",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "hdu_list = fits.open('/data/astro/scratch/lcabayol/Euclid/NNphotozs/euclid_cosmos_DC2_S2_v2.1_full.fits')\n",
+    "cat_all = Table(hdu_list[1].data).to_pandas()\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 23,
+   "id": "77924201-557e-4174-b040-2e2c6e309fc7",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>ID</th>\n",
+       "      <th>RA</th>\n",
+       "      <th>DEC</th>\n",
+       "      <th>FLUX_G_1</th>\n",
+       "      <th>FLUX_G_2</th>\n",
+       "      <th>FLUX_G_3</th>\n",
+       "      <th>FLUX_R_1</th>\n",
+       "      <th>FLUX_R_2</th>\n",
+       "      <th>FLUX_R_3</th>\n",
+       "      <th>FLUX_I_1</th>\n",
+       "      <th>...</th>\n",
+       "      <th>mu_class_L07</th>\n",
+       "      <th>photo_z_L15</th>\n",
+       "      <th>z_spec_S15</th>\n",
+       "      <th>Q_f_S15</th>\n",
+       "      <th>Instr_S15</th>\n",
+       "      <th>reliable_S15</th>\n",
+       "      <th>flag_X_ray_s15</th>\n",
+       "      <th>flag_IRAC_s15</th>\n",
+       "      <th>STAR</th>\n",
+       "      <th>AGN</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>391281</th>\n",
+       "      <td>391282</td>\n",
+       "      <td>149.553146</td>\n",
+       "      <td>2.735168</td>\n",
+       "      <td>3.187</td>\n",
+       "      <td>3.208</td>\n",
+       "      <td>3.167</td>\n",
+       "      <td>8.187</td>\n",
+       "      <td>8.348</td>\n",
+       "      <td>8.272</td>\n",
+       "      <td>12.89</td>\n",
+       "      <td>...</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0.322</td>\n",
+       "      <td>0.299501</td>\n",
+       "      <td>2.0</td>\n",
+       "      <td>PRIMUS</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>1 rows × 123 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "            ID          RA       DEC  FLUX_G_1  FLUX_G_2  FLUX_G_3  FLUX_R_1  \\\n",
+       "391281  391282  149.553146  2.735168     3.187     3.208     3.167     8.187   \n",
+       "\n",
+       "        FLUX_R_2  FLUX_R_3  FLUX_I_1  ...  mu_class_L07  photo_z_L15  \\\n",
+       "391281     8.348     8.272     12.89  ...           -99        0.322   \n",
+       "\n",
+       "        z_spec_S15  Q_f_S15   Instr_S15  reliable_S15  flag_X_ray_s15  \\\n",
+       "391281    0.299501      2.0  PRIMUS               -99               0   \n",
+       "\n",
+       "        flag_IRAC_s15  STAR  AGN  \n",
+       "391281              0     0    0  \n",
+       "\n",
+       "[1 rows x 123 columns]"
+      ]
+     },
+     "execution_count": 23,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "cat_all[(cat_all.RA==149.553146)&(cat_all.DEC==2.735168)]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 24,
+   "id": "34596af4-352c-4e7b-bce9-0724a67c5edb",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "df_ra_dec = cat_all[['RA', 'DEC']].values\n",
+    "ra_dec_pairs = cat_valid[['RA', 'DEC']].values"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 25,
+   "id": "b7c1f725-b0b3-4480-ab69-7b811727b027",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "# Find the common rows using numpy's isin() function\n",
+    "common_rows = np.isin(df_ra_dec, ra_dec_pairs).all(axis=1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 27,
+   "id": "ce02f1f7-4cf8-49d1-81ef-0b1c90338e12",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "# Filter the dataframe based on matching RA, DEC pairs\n",
+    "cat_valid_match = cat_all[common_rows]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 38,
+   "id": "4891c5be-d1f9-4dbc-84e2-c05638ce5182",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>ID</th>\n",
+       "      <th>RA</th>\n",
+       "      <th>DEC</th>\n",
+       "      <th>FLUX_G_1</th>\n",
+       "      <th>FLUX_G_2</th>\n",
+       "      <th>FLUX_G_3</th>\n",
+       "      <th>FLUX_R_1</th>\n",
+       "      <th>FLUX_R_2</th>\n",
+       "      <th>FLUX_R_3</th>\n",
+       "      <th>FLUX_I_1</th>\n",
+       "      <th>...</th>\n",
+       "      <th>mu_class_L07</th>\n",
+       "      <th>photo_z_L15</th>\n",
+       "      <th>z_spec_S15</th>\n",
+       "      <th>Q_f_S15</th>\n",
+       "      <th>Instr_S15</th>\n",
+       "      <th>reliable_S15</th>\n",
+       "      <th>flag_X_ray_s15</th>\n",
+       "      <th>flag_IRAC_s15</th>\n",
+       "      <th>STAR</th>\n",
+       "      <th>AGN</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>368</th>\n",
+       "      <td>369</td>\n",
+       "      <td>149.862381</td>\n",
+       "      <td>1.624455</td>\n",
+       "      <td>0.4753</td>\n",
+       "      <td>0.4252</td>\n",
+       "      <td>0.4659</td>\n",
+       "      <td>1.507</td>\n",
+       "      <td>1.374</td>\n",
+       "      <td>1.2680</td>\n",
+       "      <td>3.983</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1</td>\n",
+       "      <td>1.189000</td>\n",
+       "      <td>1.279900</td>\n",
+       "      <td>1.5</td>\n",
+       "      <td>zBRIGHT</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>614</th>\n",
+       "      <td>615</td>\n",
+       "      <td>149.967209</td>\n",
+       "      <td>1.625431</td>\n",
+       "      <td>1.3970</td>\n",
+       "      <td>1.3210</td>\n",
+       "      <td>1.2990</td>\n",
+       "      <td>3.310</td>\n",
+       "      <td>3.165</td>\n",
+       "      <td>2.8200</td>\n",
+       "      <td>3.815</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0.499900</td>\n",
+       "      <td>0.498300</td>\n",
+       "      <td>2.5</td>\n",
+       "      <td>zBRIGHT</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>863</th>\n",
+       "      <td>864</td>\n",
+       "      <td>150.029968</td>\n",
+       "      <td>1.625488</td>\n",
+       "      <td>1.8440</td>\n",
+       "      <td>1.5770</td>\n",
+       "      <td>1.5760</td>\n",
+       "      <td>3.141</td>\n",
+       "      <td>3.037</td>\n",
+       "      <td>2.9770</td>\n",
+       "      <td>6.058</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0.818700</td>\n",
+       "      <td>0.838300</td>\n",
+       "      <td>2.5</td>\n",
+       "      <td>zBRIGHT</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>951</th>\n",
+       "      <td>952</td>\n",
+       "      <td>149.713226</td>\n",
+       "      <td>1.625920</td>\n",
+       "      <td>2.3280</td>\n",
+       "      <td>2.3360</td>\n",
+       "      <td>2.3340</td>\n",
+       "      <td>5.336</td>\n",
+       "      <td>5.136</td>\n",
+       "      <td>5.1460</td>\n",
+       "      <td>8.928</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0.611600</td>\n",
+       "      <td>0.615000</td>\n",
+       "      <td>1.5</td>\n",
+       "      <td>zBRIGHT</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1292</th>\n",
+       "      <td>1293</td>\n",
+       "      <td>149.530899</td>\n",
+       "      <td>1.626842</td>\n",
+       "      <td>1.0330</td>\n",
+       "      <td>0.8639</td>\n",
+       "      <td>0.8805</td>\n",
+       "      <td>1.963</td>\n",
+       "      <td>1.939</td>\n",
+       "      <td>1.9980</td>\n",
+       "      <td>3.388</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0.751600</td>\n",
+       "      <td>0.763700</td>\n",
+       "      <td>9.5</td>\n",
+       "      <td>zBRIGHT</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>391055</th>\n",
+       "      <td>391056</td>\n",
+       "      <td>149.911942</td>\n",
+       "      <td>2.737337</td>\n",
+       "      <td>2.5840</td>\n",
+       "      <td>2.7920</td>\n",
+       "      <td>2.7740</td>\n",
+       "      <td>4.309</td>\n",
+       "      <td>4.247</td>\n",
+       "      <td>4.0800</td>\n",
+       "      <td>6.144</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0.207600</td>\n",
+       "      <td>9.999900</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>zBRIGHT</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>391111</th>\n",
+       "      <td>391112</td>\n",
+       "      <td>149.960968</td>\n",
+       "      <td>2.735883</td>\n",
+       "      <td>1.1160</td>\n",
+       "      <td>1.0270</td>\n",
+       "      <td>1.0280</td>\n",
+       "      <td>4.790</td>\n",
+       "      <td>4.807</td>\n",
+       "      <td>4.8790</td>\n",
+       "      <td>14.450</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0.690500</td>\n",
+       "      <td>0.696300</td>\n",
+       "      <td>1.5</td>\n",
+       "      <td>zBRIGHT</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>391133</th>\n",
+       "      <td>391134</td>\n",
+       "      <td>149.802002</td>\n",
+       "      <td>2.735837</td>\n",
+       "      <td>1.1510</td>\n",
+       "      <td>0.9859</td>\n",
+       "      <td>1.0060</td>\n",
+       "      <td>2.038</td>\n",
+       "      <td>2.059</td>\n",
+       "      <td>1.9160</td>\n",
+       "      <td>3.724</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1</td>\n",
+       "      <td>-99.900002</td>\n",
+       "      <td>0.680074</td>\n",
+       "      <td>2.0</td>\n",
+       "      <td>PRIMUS</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>391178</th>\n",
+       "      <td>391179</td>\n",
+       "      <td>149.902161</td>\n",
+       "      <td>2.735792</td>\n",
+       "      <td>0.6982</td>\n",
+       "      <td>0.6542</td>\n",
+       "      <td>0.5222</td>\n",
+       "      <td>1.162</td>\n",
+       "      <td>1.024</td>\n",
+       "      <td>0.7999</td>\n",
+       "      <td>3.115</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1</td>\n",
+       "      <td>0.860600</td>\n",
+       "      <td>0.881794</td>\n",
+       "      <td>2.0</td>\n",
+       "      <td>PRIMUS</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>391281</th>\n",
+       "      <td>391282</td>\n",
+       "      <td>149.553146</td>\n",
+       "      <td>2.735168</td>\n",
+       "      <td>3.1870</td>\n",
+       "      <td>3.2080</td>\n",
+       "      <td>3.1670</td>\n",
+       "      <td>8.187</td>\n",
+       "      <td>8.348</td>\n",
+       "      <td>8.2720</td>\n",
+       "      <td>12.890</td>\n",
+       "      <td>...</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0.322000</td>\n",
+       "      <td>0.299501</td>\n",
+       "      <td>2.0</td>\n",
+       "      <td>PRIMUS</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>10057 rows × 123 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "            ID          RA       DEC  FLUX_G_1  FLUX_G_2  FLUX_G_3  FLUX_R_1  \\\n",
+       "368        369  149.862381  1.624455    0.4753    0.4252    0.4659     1.507   \n",
+       "614        615  149.967209  1.625431    1.3970    1.3210    1.2990     3.310   \n",
+       "863        864  150.029968  1.625488    1.8440    1.5770    1.5760     3.141   \n",
+       "951        952  149.713226  1.625920    2.3280    2.3360    2.3340     5.336   \n",
+       "1292      1293  149.530899  1.626842    1.0330    0.8639    0.8805     1.963   \n",
+       "...        ...         ...       ...       ...       ...       ...       ...   \n",
+       "391055  391056  149.911942  2.737337    2.5840    2.7920    2.7740     4.309   \n",
+       "391111  391112  149.960968  2.735883    1.1160    1.0270    1.0280     4.790   \n",
+       "391133  391134  149.802002  2.735837    1.1510    0.9859    1.0060     2.038   \n",
+       "391178  391179  149.902161  2.735792    0.6982    0.6542    0.5222     1.162   \n",
+       "391281  391282  149.553146  2.735168    3.1870    3.2080    3.1670     8.187   \n",
+       "\n",
+       "        FLUX_R_2  FLUX_R_3  FLUX_I_1  ...  mu_class_L07  photo_z_L15  \\\n",
+       "368        1.374    1.2680     3.983  ...             1     1.189000   \n",
+       "614        3.165    2.8200     3.815  ...             1     0.499900   \n",
+       "863        3.037    2.9770     6.058  ...             1     0.818700   \n",
+       "951        5.136    5.1460     8.928  ...             1     0.611600   \n",
+       "1292       1.939    1.9980     3.388  ...             1     0.751600   \n",
+       "...          ...       ...       ...  ...           ...          ...   \n",
+       "391055     4.247    4.0800     6.144  ...             1     0.207600   \n",
+       "391111     4.807    4.8790    14.450  ...             1     0.690500   \n",
+       "391133     2.059    1.9160     3.724  ...             1   -99.900002   \n",
+       "391178     1.024    0.7999     3.115  ...             1     0.860600   \n",
+       "391281     8.348    8.2720    12.890  ...           -99     0.322000   \n",
+       "\n",
+       "        z_spec_S15  Q_f_S15   Instr_S15  reliable_S15  flag_X_ray_s15  \\\n",
+       "368       1.279900      1.5  zBRIGHT              -99               0   \n",
+       "614       0.498300      2.5  zBRIGHT              -99               0   \n",
+       "863       0.838300      2.5  zBRIGHT              -99               0   \n",
+       "951       0.615000      1.5  zBRIGHT              -99               0   \n",
+       "1292      0.763700      9.5  zBRIGHT              -99               0   \n",
+       "...            ...      ...         ...           ...             ...   \n",
+       "391055    9.999900      0.0  zBRIGHT              -99               0   \n",
+       "391111    0.696300      1.5  zBRIGHT              -99               0   \n",
+       "391133    0.680074      2.0  PRIMUS               -99               0   \n",
+       "391178    0.881794      2.0  PRIMUS               -99               0   \n",
+       "391281    0.299501      2.0  PRIMUS               -99               0   \n",
+       "\n",
+       "        flag_IRAC_s15  STAR  AGN  \n",
+       "368                 0     0    0  \n",
+       "614                 0     0    0  \n",
+       "863                 0     0    0  \n",
+       "951                 0     0    0  \n",
+       "1292                0     0    0  \n",
+       "...               ...   ...  ...  \n",
+       "391055              0     0    0  \n",
+       "391111              0     0    0  \n",
+       "391133              0     0    0  \n",
+       "391178              0     0    0  \n",
+       "391281              0     0    0  \n",
+       "\n",
+       "[10057 rows x 123 columns]"
+      ]
+     },
+     "execution_count": 38,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "cat_valid_match[(cat_valid_match.reliable_S15<0)&(cat_valid_match.z_spec_S15>0)]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 33,
+   "id": "2d4324c9-2f8d-4aac-9435-c25b91cb4b26",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "cat_valid_match_fitTable = Table.from_pandas(cat_valid_match)\n",
+    "\n",
+    "# Define the output file path\n",
+    "output_file_path = \"/data/astro/scratch/lcabayol/Euclid/NNphotozs/Euclid_EXT_MER_PHZ_DC2_v1.5/euclid_cosmos_DC2_S1_v2.1_valid_matched.fits\"\n",
+    "\n",
+    "# Save the FITS table to the output file\n",
+    "cat_valid_match_fitTable.write(output_file_path, format='fits', overwrite=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "8b8b25f3-1f2b-4e61-b644-c0be213d9449",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 72,
+   "id": "7df1acba-15a2-4555-b2ae-6ce4ad55df4b",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([  0. ,   3. ,   9. ,   1. ,   4. ,  20. ,   9.5,   2. , -10. ,\n",
+       "        22. ,  21. ,   2.5,   3.5,   3.1,  14. ,  13.1,  10. ,   1.1,\n",
+       "        13. ,  32. ,  13.5,   1.5,  33. ,   2.1,   4.5,   4.1,  14.5,\n",
+       "        29. ,  39. ,  22.5], dtype=float32)"
+      ]
+     },
+     "execution_count": 72,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "cat_calib[cat_calib.z_spec_S15>3].Q_f_S15.unique()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 66,
+   "id": "13ab99c4-0c15-46d2-8c57-27eeb1d6c762",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([-99. ,   2.5,   3.5,   4.5,   0. ,  21.1,   4. ,   1.5,  13.5,\n",
+       "         1.1,  23.5,   9. ,   3. ,  24.4,   2. ,   2.1,   4.1,   1. ,\n",
+       "         9.5,   3.1,  23.1,  29.5,   4.4,   9.1,  22.4,  22.5,   2.4,\n",
+       "         9.3,  24.5,  22.1,  20. ,  14.5,  11.5,   6. , -10. ,  22. ,\n",
+       "        23. ,  21. ,  24. ,  31. ,   3.4,   1.4,  -1. ,  13.1,  18.1,\n",
+       "        21.5,  29.4,  12.1,  39. ,  14. ,  23.4,  29. ,  19. ,   0.5,\n",
+       "        12.5,  29.3,  10. ,  13. ,  24.1,  34. ,  14.1,  32. ,  21.4,\n",
+       "        33. ,  18.5,  29.1,   5. ,  18.3,  11.1,  14.4,  12. ,   9.4,\n",
+       "         5.1], dtype=float32)"
+      ]
+     },
+     "execution_count": 66,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "cat_calib.Q_f_S15.unique()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 94,
+   "id": "12c171b0-8465-47e0-b894-243837e02795",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "weight_dict={(-99,0.99):0,\n",
+    "             (1,1.99):0.5,\n",
+    "             (2,2.99):0.75,\n",
+    "             (3,4):1,\n",
+    "             (9,9.99):0.25,\n",
+    "             (10,10.99):0,\n",
+    "             (11,11.99):0.5,\n",
+    "             (12,12.99):0.75,\n",
+    "             (13,14):1,\n",
+    "             (14.01,40):0\n",
+    "            }"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 96,
+   "id": "d3080023-1af2-46cb-8294-034d47d52a41",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "def map_weight(Qz):\n",
+    "    for key, value in weight_dict.items():\n",
+    "        if key[0] <= Qz <= key[1]:\n",
+    "            return value\n",
+    "    return None\n",
+    "\n",
+    "# Apply the function to create the 'wQz' column\n",
+    "cat_calib['w_Q_f_S15'] = cat_calib['Q_f_S15'].apply(map_weight)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 93,
+   "id": "3367b799-da93-4710-b41e-55bb2a2d4059",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>ID</th>\n",
+       "      <th>RA</th>\n",
+       "      <th>DEC</th>\n",
+       "      <th>FLUX_G_1</th>\n",
+       "      <th>FLUX_G_2</th>\n",
+       "      <th>FLUX_G_3</th>\n",
+       "      <th>FLUX_R_1</th>\n",
+       "      <th>FLUX_R_2</th>\n",
+       "      <th>FLUX_R_3</th>\n",
+       "      <th>FLUX_I_1</th>\n",
+       "      <th>...</th>\n",
+       "      <th>photo_z_L15</th>\n",
+       "      <th>z_spec_S15</th>\n",
+       "      <th>Q_f_S15</th>\n",
+       "      <th>Instr_S15</th>\n",
+       "      <th>reliable_S15</th>\n",
+       "      <th>flag_X_ray_s15</th>\n",
+       "      <th>flag_IRAC_s15</th>\n",
+       "      <th>STAR</th>\n",
+       "      <th>AGN</th>\n",
+       "      <th>wQz</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>32</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>0.27910</td>\n",
+       "      <td>0.26540</td>\n",
+       "      <td>-0.060160</td>\n",
+       "      <td>0.16420</td>\n",
+       "      <td>0.10770</td>\n",
+       "      <td>0.09359</td>\n",
+       "      <td>0.6225</td>\n",
+       "      <td>...</td>\n",
+       "      <td>2.095800</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>36</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>0.16160</td>\n",
+       "      <td>0.11760</td>\n",
+       "      <td>0.093950</td>\n",
+       "      <td>0.13680</td>\n",
+       "      <td>0.02803</td>\n",
+       "      <td>0.06321</td>\n",
+       "      <td>0.3125</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.138100</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>38</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>0.20970</td>\n",
+       "      <td>0.23170</td>\n",
+       "      <td>0.199000</td>\n",
+       "      <td>0.38770</td>\n",
+       "      <td>0.39770</td>\n",
+       "      <td>0.33170</td>\n",
+       "      <td>0.1775</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1.080600</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>39</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>0.15680</td>\n",
+       "      <td>0.04144</td>\n",
+       "      <td>0.006729</td>\n",
+       "      <td>0.32470</td>\n",
+       "      <td>0.28490</td>\n",
+       "      <td>0.10140</td>\n",
+       "      <td>0.2689</td>\n",
+       "      <td>...</td>\n",
+       "      <td>-99.000000</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>40</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>0.29370</td>\n",
+       "      <td>0.36790</td>\n",
+       "      <td>0.381100</td>\n",
+       "      <td>0.59510</td>\n",
+       "      <td>0.48770</td>\n",
+       "      <td>0.55310</td>\n",
+       "      <td>0.2876</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1.601600</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>190681</th>\n",
+       "      <td>197465</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>0.23410</td>\n",
+       "      <td>0.11830</td>\n",
+       "      <td>0.060100</td>\n",
+       "      <td>0.14460</td>\n",
+       "      <td>0.24370</td>\n",
+       "      <td>0.46160</td>\n",
+       "      <td>0.2579</td>\n",
+       "      <td>...</td>\n",
+       "      <td>-99.900002</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>190682</th>\n",
+       "      <td>197479</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>0.04739</td>\n",
+       "      <td>0.04522</td>\n",
+       "      <td>0.036710</td>\n",
+       "      <td>0.17800</td>\n",
+       "      <td>0.16930</td>\n",
+       "      <td>0.10800</td>\n",
+       "      <td>0.3385</td>\n",
+       "      <td>...</td>\n",
+       "      <td>-99.000000</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>190683</th>\n",
+       "      <td>197490</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>4.81900</td>\n",
+       "      <td>4.76700</td>\n",
+       "      <td>4.774000</td>\n",
+       "      <td>12.73000</td>\n",
+       "      <td>12.71000</td>\n",
+       "      <td>12.67000</td>\n",
+       "      <td>20.5300</td>\n",
+       "      <td>...</td>\n",
+       "      <td>-99.900002</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>190684</th>\n",
+       "      <td>197492</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-0.16100</td>\n",
+       "      <td>-0.48150</td>\n",
+       "      <td>-0.490300</td>\n",
+       "      <td>0.29440</td>\n",
+       "      <td>-0.63920</td>\n",
+       "      <td>-0.05621</td>\n",
+       "      <td>-1.5310</td>\n",
+       "      <td>...</td>\n",
+       "      <td>-99.000000</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>190685</th>\n",
+       "      <td>197497</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>0.10890</td>\n",
+       "      <td>0.07218</td>\n",
+       "      <td>0.086700</td>\n",
+       "      <td>0.00439</td>\n",
+       "      <td>0.06940</td>\n",
+       "      <td>0.07060</td>\n",
+       "      <td>0.3944</td>\n",
+       "      <td>...</td>\n",
+       "      <td>-99.900002</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99.0</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>-99</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>190686 rows × 124 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "            ID    RA   DEC  FLUX_G_1  FLUX_G_2  FLUX_G_3  FLUX_R_1  FLUX_R_2  \\\n",
+       "0           32 -99.0 -99.0   0.27910   0.26540 -0.060160   0.16420   0.10770   \n",
+       "1           36 -99.0 -99.0   0.16160   0.11760  0.093950   0.13680   0.02803   \n",
+       "2           38 -99.0 -99.0   0.20970   0.23170  0.199000   0.38770   0.39770   \n",
+       "3           39 -99.0 -99.0   0.15680   0.04144  0.006729   0.32470   0.28490   \n",
+       "4           40 -99.0 -99.0   0.29370   0.36790  0.381100   0.59510   0.48770   \n",
+       "...        ...   ...   ...       ...       ...       ...       ...       ...   \n",
+       "190681  197465 -99.0 -99.0   0.23410   0.11830  0.060100   0.14460   0.24370   \n",
+       "190682  197479 -99.0 -99.0   0.04739   0.04522  0.036710   0.17800   0.16930   \n",
+       "190683  197490 -99.0 -99.0   4.81900   4.76700  4.774000  12.73000  12.71000   \n",
+       "190684  197492 -99.0 -99.0  -0.16100  -0.48150 -0.490300   0.29440  -0.63920   \n",
+       "190685  197497 -99.0 -99.0   0.10890   0.07218  0.086700   0.00439   0.06940   \n",
+       "\n",
+       "        FLUX_R_3  FLUX_I_1  ...  photo_z_L15  z_spec_S15  Q_f_S15  Instr_S15  \\\n",
+       "0        0.09359    0.6225  ...     2.095800       -99.0    -99.0        -99   \n",
+       "1        0.06321    0.3125  ...     0.138100       -99.0    -99.0        -99   \n",
+       "2        0.33170    0.1775  ...     1.080600       -99.0    -99.0        -99   \n",
+       "3        0.10140    0.2689  ...   -99.000000       -99.0    -99.0        -99   \n",
+       "4        0.55310    0.2876  ...     1.601600       -99.0    -99.0        -99   \n",
+       "...          ...       ...  ...          ...         ...      ...        ...   \n",
+       "190681   0.46160    0.2579  ...   -99.900002       -99.0    -99.0        -99   \n",
+       "190682   0.10800    0.3385  ...   -99.000000       -99.0    -99.0        -99   \n",
+       "190683  12.67000   20.5300  ...   -99.900002       -99.0    -99.0        -99   \n",
+       "190684  -0.05621   -1.5310  ...   -99.000000       -99.0    -99.0        -99   \n",
+       "190685   0.07060    0.3944  ...   -99.900002       -99.0    -99.0        -99   \n",
+       "\n",
+       "        reliable_S15  flag_X_ray_s15  flag_IRAC_s15  STAR  AGN  wQz  \n",
+       "0                -99               0              0     0    0  NaN  \n",
+       "1                -99               0              0     0    0  NaN  \n",
+       "2                -99               0              0     0    0  NaN  \n",
+       "3                -99               0              0     0    0  NaN  \n",
+       "4                -99               0              0     0    0  NaN  \n",
+       "...              ...             ...            ...   ...  ...  ...  \n",
+       "190681           -99               0              0     0    0  NaN  \n",
+       "190682           -99               0              0     0    0  NaN  \n",
+       "190683           -99               0              0     0    0  NaN  \n",
+       "190684           -99               0              0     0    0  NaN  \n",
+       "190685           -99               0              0     0    0  NaN  \n",
+       "\n",
+       "[190686 rows x 124 columns]"
+      ]
+     },
+     "execution_count": 93,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "cat_calib"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "968017d9-c4d7-4891-b74a-4422ee7bb73c",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "DLenv2",
+   "language": "python",
+   "name": "dlenv2"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.7"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}