NIST-In-Situ-IN625-LPBF-Overhangs_old.py

import datasets
import numpy as np
import os
import pickle
import pyarrow.parquet as pq

_DESCRIPTION = """\
In Situ Thermography During Laser Powder Bed Fusion of a Nickel Superalloy 625
Artifact with Various Overhangs and Supports
"""

LAYER_BASE_URLS = []

# Layers 1 to 99 inclusive without layer 22
for layer_number in range(1, 100, 1):
    if layer_number != 22:
        LAYER_BASE_URLS.append(f"./layer/base/{layer_number}.pkl")

LAYER_OVERHANG_WITH_SUPPORTS_URLS = [f"./layer/overhang_with_supports/{n}.pkl" for n in range(101, 281, 1)]
LAYER_BLOCK_URLS = [f"./layer/block/{n}.pkl" for n in range(281, 381, 1)]
LAYER_OVERHANG_NO_SUPPORTS_URLS = [f"./layer/overhang_no_supports/{n}.pkl" for n in range(381, 560, 1)]

LAYER_TABLE_BASE_URLS = []

# Layers 1 to 99 inclusive without layer 22
for layer_number in range(1, 100, 1):
    if layer_number != 22:
        LAYER_TABLE_BASE_URLS.append(f"./layer_table/base/{layer_number}.parquet")

LAYER_TABLE_OVERHANG_WITH_SUPPORTS_URLS = [f"./layer_table/overhang_with_supports/{n}.parquet" for n in range(101, 281, 1)]
LAYER_TABLE_BLOCK_URLS = [f"./layer_table/block/{n}.parquet" for n in range(281, 381, 1)]
LAYER_TABLE_OVERHANG_NO_SUPPORTS_URLS = [f"./layer_table/overhang_no_supports/{n}.parquet" for n in range(381, 560, 1)]

_URLS = {
    "part_section": {
        "base": "./part_section/BASE.pkl",
        "block": "./part_section/BLOCK.pkl",
        "overhang_no_supports": "./part_section/OVERHANG_noSup.pkl",
        "overhang_with_supports": "./part_section/OVERHANG_wSup.pkl",
    },
    "layer": {
        "base": LAYER_BASE_URLS,
        "block": LAYER_BLOCK_URLS,
        "overhang_no_supports": LAYER_OVERHANG_NO_SUPPORTS_URLS,
        "overhang_with_supports": LAYER_OVERHANG_WITH_SUPPORTS_URLS,
    },
    "layer_table": {
        "base": LAYER_TABLE_BASE_URLS,
        "block": LAYER_TABLE_BLOCK_URLS,
        "overhang_no_supports": LAYER_TABLE_OVERHANG_NO_SUPPORTS_URLS,
        "overhang_with_supports": LAYER_TABLE_OVERHANG_WITH_SUPPORTS_URLS,
    }
}

class NISTInSituIN625LPBFOverhangsDataset(datasets.GeneratorBasedBuilder):
    VERSION = datasets.Version("0.0.1")

    BUILDER_CONFIGS = [
        datasets.BuilderConfig(
            name="part_section",
            description="Original dataset files split by part section",
            version=VERSION,
        ),
        datasets.BuilderConfig(
            name="layer",
            description="Provides layer-wise attributes of entire dataset",
            version=VERSION,
        ),
        datasets.BuilderConfig(
            name="layer_table",
            description="Provides parquet layer-wise attributes of entire dataset",
            version=VERSION,
        ),
    ]

    DEFAULT_CONFIG_NAME = "layer_table"

    def _info(self):
        features = datasets.Features({
            "folder_layer_range": datasets.Value("string"),
            "part": datasets.Value("string"),
            "part_section": datasets.Value("string"),
            "process": datasets.Value("string"),
            "source": datasets.Value("string"),
            "layer_number": datasets.Value("string"),
            "build_time": datasets.Sequence(datasets.Sequence(datasets.Value("float32"))),
            "contact_email": datasets.Value("string"),
            "file_name": datasets.Value("string"),
            "hatch_spacing": datasets.Value("uint32"),
            "laser_power": datasets.Value("uint32"),
            "layer_thickness": datasets.Value("uint32"),
            "material": datasets.Value("string"),
            "radiant_temp": datasets.Sequence(datasets.Sequence(datasets.Sequence(datasets.Value("uint32")))),
            "build_time": datasets.Sequence(datasets.Sequence(datasets.Value("float32"))),
            "s_hvariable__a": datasets.Value("float32"),
            "s_hvariable__b": datasets.Value("float32"),
            "s_hvariable__c": datasets.Value("float32"),
            "scan_speed": datasets.Value("uint32"),
            "website": datasets.Value("string"),
        })

        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,
        )

    def _split_generators(self, dl_manager):

        urls = _URLS[self.config.name]
        downloaded_files = dl_manager.download_and_extract(urls)

        return [
            datasets.SplitGenerator(
                name="base",
                gen_kwargs={
                    "filepath": downloaded_files["base"],
                    "split": "base",
                }
            ),
            datasets.SplitGenerator(
                name="block",
                gen_kwargs={
                    "filepath": downloaded_files["block"],
                    "split": "block",
                }
            ),
            datasets.SplitGenerator(
                name="overhang_no_supports",
                gen_kwargs={
                    "filepath": downloaded_files["overhang_no_supports"],
                    "split": "overhang_no_supports",
                }
            ),
            datasets.SplitGenerator(
                name="overhang_with_supports",
                gen_kwargs={
                    "filepath": downloaded_files["overhang_with_supports"],
                    "split": "overhang_with_supports",
                }
            ),
        ]
    
    def _generate_examples(self, filepath, split):
        if self.config.name == "part_section":
            with open(filepath, "rb") as f:
                layers = pickle.load(f)
                for index, layer in enumerate(layers):
                    yield index, layer
        elif self.config.name == "layer":
            # layer config has multiple files in filepath variable.
            for index, path in enumerate(filepath):
                with open(path, "rb") as f:
                    layer = pickle.load(f)
                    yield index, layer
        elif self.config.name == "layer_table":
            # layer config has multiple files in filepath variable.
            for index, path in enumerate(filepath):
                with open(path, "rb") as f:
                    table = pq.read_table(f)
                    layer = table.to_pydict()
                    non_array = [str, int, float]
                    converted_layer = {}
                    for key, value in layer.items():
                        layer_value = value[0]
                        if (type(layer_value) in non_array):
                            converted_layer[key] = layer_value
                        elif(isinstance(value, list) and "shape" not in key):
                            shape = layer[f"{key}_shape"][0]
                            flattened_array = np.array(layer_value)
                            array = flattened_array.reshape(shape)
                            converted_layer[key] = array

                    yield index, converted_layer