Create deepbindweight

deepbindweight

@@ -0,0 +1,52 @@
+import os
+import json
+import re
+import pandas as pd
+import datasets
+
+
+logger = datasets.logging.get_logger(__name__)
+
+ID_POOL = ()
+URL = "https://huggingface.co/datasets/thewall/DeepBindWeight/blob/main/params"
+
+class DeepBindConfig(datasets.BuilderConfig):
+    pass
+
+
+class JolmaLocal(datasets.GeneratorBasedBuilder):
+    BUILDER_CONFIGS = [
+        DeepBindConfig(name=key) for key in ID_POOL
+    ]
+
+    DEFAULT_CONFIG_NAME = "D00328.003"
+
+    def _info(self):
+        return datasets.DatasetInfo(
+            features=datasets.Features(
+                {
+                    "config": datasets.Value("string"),
+                    "existed": datasets.Value("bool")
+                }
+            ),
+            homepage="http://tools.genes.toronto.edu/deepbind",
+        )
+
+    def _split_generators(self, dl_manager):
+        url = f"{URL}/{self.config.name}.txt"
+        downloaded_files = dl_manager.download_and_extract(url)
+        return [
+            datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files}),
+        ]
+
+    def _generate_examples(self, filepath):
+        """This function returns the examples in the raw (text) form."""
+        logger.info("generating examples from = %s", filepath)
+        yield 0, {"config": filepath, "existed": os.path.exists(filepath)}
+
+
+if __name__=="__main__":
+    from datasets import load_dataset
+    dataset = load_dataset("deepbindweight.py", split="all")
+    dataset.push_to_hub("thewall/DeepBindWeight")
+