Dea22 commited on
Commit
e6f71f0
1 Parent(s): 7b8fa75

added progress

Browse files
Files changed (1) hide show
  1. app.py +25 -17
app.py CHANGED
@@ -2,25 +2,31 @@ import numpy as np
2
  import matplotlib.pyplot as plt
3
  from sklearn.linear_model import MultiTaskLasso, Lasso
4
  import gradio as gr
 
5
 
6
  rng = np.random.RandomState(42)
7
 
8
  # Generate some 2D coefficients with sine waves with random frequency and phase
9
- def make_plot(n_samples, n_features, n_tasks, n_relevant_features, alpha):
 
 
 
 
 
10
 
11
  coef = np.zeros((n_tasks, n_features))
12
  times = np.linspace(0, 2 * np.pi, n_tasks)
13
  for k in range(n_relevant_features):
14
  coef[:, k] = np.sin((1.0 + rng.randn(1)) * times + 3 * rng.randn(1))
15
-
16
  X = rng.randn(n_samples, n_features)
17
  Y = np.dot(X, coef.T) + rng.randn(n_samples, n_tasks)
18
-
19
  coef_lasso_ = np.array([Lasso(alpha=0.5).fit(X, y).coef_ for y in Y.T])
20
  coef_multi_task_lasso_ = MultiTaskLasso(alpha=alpha).fit(X, Y).coef_
21
-
22
  fig = plt.figure(figsize=(8, 5))
23
-
24
  feature_to_plot = 0
25
  fig = plt.figure()
26
  lw = 2
@@ -34,14 +40,15 @@ def make_plot(n_samples, n_features, n_tasks, n_relevant_features, alpha):
34
  linewidth=lw,
35
  label="MultiTaskLasso",
36
  )
 
37
  plt.legend(loc='upper center', bbox_to_anchor=(0.5, -0.05),
38
  ncol=3, fancybox=True, shadow=True)
39
  plt.axis("tight")
40
  plt.ylim([-1.1, 1.1])
41
  fig.suptitle("Lasso, MultiTaskLasso and Ground truth time series")
42
  return fig
43
-
44
-
45
  model_card=f"""
46
  ## Description
47
  Multi-task Lasso allows us to jointly fit multiple regression problems by enforcing the selected
@@ -56,7 +63,7 @@ Plots represent Lasso, MultiTaskLasso and Ground truth time series
56
  """
57
 
58
  with gr.Blocks() as demo:
59
-
60
  gr.Markdown('''
61
  <div>
62
  <h1 style='text-align: center'> Joint feature selection with multi-task Lasso </h1>
@@ -67,16 +74,17 @@ with gr.Blocks() as demo:
67
  gr.Markdown(
68
  "Iterative conversion by: <a href=\"https://www.deamarialeon.com\">Dea María Léon</a>"
69
  )
70
- n_samples = gr.Slider(50,500,value=100,step=50,label='Select number of samples')
71
- n_features = gr.Slider(5,50,value=30,step=5,label='Select number of features')
72
- n_tasks = gr.Slider(5,50,value=40,step=5,label='Select number of tasks')
73
- n_relevant_features = gr.Slider(1,10,value=5,step=1,label='Select number of relevant_features')
74
- with gr.Column():
75
- with gr.Tab('Select Alpha Range'):
76
- alpha = gr.Slider(0,10,value=1.0,step=0.5,label='alpha')
77
-
 
78
  btn = gr.Button(value = 'Submit')
79
 
80
  btn.click(make_plot,inputs=[n_samples,n_features, n_tasks, n_relevant_features, alpha],outputs=[gr.Plot()])
81
 
82
- demo.launch()
 
2
  import matplotlib.pyplot as plt
3
  from sklearn.linear_model import MultiTaskLasso, Lasso
4
  import gradio as gr
5
+ import time
6
 
7
  rng = np.random.RandomState(42)
8
 
9
  # Generate some 2D coefficients with sine waves with random frequency and phase
10
+ def make_plot(n_samples, n_features, n_tasks, n_relevant_features, alpha, progress=gr.Progress()):
11
+
12
+ progress(0, desc="Starting...")
13
+ time.sleep(1)
14
+ for i in progress.tqdm(range(100)):
15
+ time.sleep(0.1)
16
 
17
  coef = np.zeros((n_tasks, n_features))
18
  times = np.linspace(0, 2 * np.pi, n_tasks)
19
  for k in range(n_relevant_features):
20
  coef[:, k] = np.sin((1.0 + rng.randn(1)) * times + 3 * rng.randn(1))
21
+
22
  X = rng.randn(n_samples, n_features)
23
  Y = np.dot(X, coef.T) + rng.randn(n_samples, n_tasks)
24
+
25
  coef_lasso_ = np.array([Lasso(alpha=0.5).fit(X, y).coef_ for y in Y.T])
26
  coef_multi_task_lasso_ = MultiTaskLasso(alpha=alpha).fit(X, Y).coef_
27
+
28
  fig = plt.figure(figsize=(8, 5))
29
+
30
  feature_to_plot = 0
31
  fig = plt.figure()
32
  lw = 2
 
40
  linewidth=lw,
41
  label="MultiTaskLasso",
42
  )
43
+ #plt.legend(loc="upper center")
44
  plt.legend(loc='upper center', bbox_to_anchor=(0.5, -0.05),
45
  ncol=3, fancybox=True, shadow=True)
46
  plt.axis("tight")
47
  plt.ylim([-1.1, 1.1])
48
  fig.suptitle("Lasso, MultiTaskLasso and Ground truth time series")
49
  return fig
50
+
51
+
52
  model_card=f"""
53
  ## Description
54
  Multi-task Lasso allows us to jointly fit multiple regression problems by enforcing the selected
 
63
  """
64
 
65
  with gr.Blocks() as demo:
66
+
67
  gr.Markdown('''
68
  <div>
69
  <h1 style='text-align: center'> Joint feature selection with multi-task Lasso </h1>
 
74
  gr.Markdown(
75
  "Iterative conversion by: <a href=\"https://www.deamarialeon.com\">Dea María Léon</a>"
76
  )
77
+
78
+ with gr.Row().style(equal_height=True):
79
+
80
+ n_features = gr.Slider(5,50,value=30,step=5,label='Features')
81
+ n_tasks = gr.Slider(5,50,value=40,step=5,label='Tasks')
82
+ n_relevant_features = gr.Slider(1,10,value=5,step=1,label='Relevant features')
83
+ alpha = gr.Slider(0,10,value=1.0,step=0.5,label='Alpha Range')
84
+ n_samples = gr.Slider(50,500,value=100,step=50,label='Number of samples')
85
+
86
  btn = gr.Button(value = 'Submit')
87
 
88
  btn.click(make_plot,inputs=[n_samples,n_features, n_tasks, n_relevant_features, alpha],outputs=[gr.Plot()])
89
 
90
+ demo.queue().launch()