Upload permutation-groups.py with huggingface_hub

permutation-groups.py

@@ -1,20 +1,18 @@
 import datasets
 import json
 import os
-import glob
-from sympy.combinatorics import Permutation
-from sympy.combinatorics.named_groups import AlternatingGroup, SymmetricGroup
 
 _DESCRIPTION = "A collection of permutation composition datasets for various symmetric and alternating groups."
 _HOMEPAGE = "https://huggingface.co/datasets/BeeGass/permutation-groups"
 _LICENSE = "MIT"
 
 class PermutationGroupsConfig(datasets.BuilderConfig):
-    def __init__(self, *args, group_name=None, group_degree=None, group_order=None, **kwargs):
+    def __init__(self, *args, group_name=None, group_degree=None, group_order=None, data_dir=None, **kwargs):
         super().__init__(*args, **kwargs)
         self.group_name = group_name
         self.group_degree = group_degree
         self.group_order = group_order
+        self.data_dir = data_dir
 
 class PermutationGroups(datasets.GeneratorBasedBuilder):
     VERSION = datasets.Version("1.0.0")
@@ -26,6 +24,7 @@ class PermutationGroups(datasets.GeneratorBasedBuilder):
             group_name="S3",
             group_degree=3,
             group_order=6,
+            data_dir="data/s3_data",
         ),
         PermutationGroupsConfig(
             name="s4_data",
@@ -33,6 +32,7 @@ class PermutationGroups(datasets.GeneratorBasedBuilder):
             group_name="S4",
             group_degree=4,
             group_order=24,
+            data_dir="data/s4_data",
         ),
         PermutationGroupsConfig(
             name="s5_data",
@@ -40,6 +40,7 @@ class PermutationGroups(datasets.GeneratorBasedBuilder):
             group_name="S5",
             group_degree=5,
             group_order=120,
+            data_dir="data/s5_data",
         ),
         PermutationGroupsConfig(
             name="s6_data",
@@ -47,6 +48,7 @@ class PermutationGroups(datasets.GeneratorBasedBuilder):
             group_name="S6",
             group_degree=6,
             group_order=720,
+            data_dir="data/s6_data",
         ),
         PermutationGroupsConfig(
             name="s7_data",
@@ -54,6 +56,7 @@ class PermutationGroups(datasets.GeneratorBasedBuilder):
             group_name="S7",
             group_degree=7,
             group_order=5040,
+            data_dir="data/s7_data",
         ),
         PermutationGroupsConfig(
             name="a5_data",
@@ -61,6 +64,7 @@ class PermutationGroups(datasets.GeneratorBasedBuilder):
             group_name="A5",
             group_degree=5,
             group_order=60,
+            data_dir="data/a5_data",
         ),
         PermutationGroupsConfig(
             name="a6_data",
@@ -68,6 +72,7 @@ class PermutationGroups(datasets.GeneratorBasedBuilder):
             group_name="A6",
             group_degree=6,
             group_order=360,
+            data_dir="data/a6_data",
         ),
         PermutationGroupsConfig(
             name="a7_data",
@@ -75,6 +80,7 @@ class PermutationGroups(datasets.GeneratorBasedBuilder):
             group_name="A7",
             group_degree=7,
             group_order=2520,
+            data_dir="data/a7_data",
         ),
         PermutationGroupsConfig(
             name="all",
@@ -82,6 +88,7 @@ class PermutationGroups(datasets.GeneratorBasedBuilder):
             group_name="All",
             group_degree=None,
             group_order=None,
+            data_dir=None,  # Special handling for 'all'
         ),
     ]
 
@@ -105,70 +112,89 @@ class PermutationGroups(datasets.GeneratorBasedBuilder):
             all_configs = ["s3_data", "s4_data", "s5_data", "s6_data", "s7_data", 
                           "a5_data", "a6_data", "a7_data"]
             
-            # For "all" config, we need to download each dataset's arrow files
-            # Use glob pattern to match any arrow files in the directories
-            data_files = {
-                "train": [f"data/{config}/train/data-*-of-*.arrow" for config in all_configs],
-                "test": [f"data/{config}/test/data-*-of-*.arrow" for config in all_configs],
-            }
+            # Download all arrow files
+            train_files = []
+            test_files = []
+            
+            for config in all_configs:
+                data_urls = {
+                    "train": f"data/{config}/train/data-00000-of-00001.arrow",
+                    "test": f"data/{config}/test/data-00000-of-00001.arrow",
+                }
+                downloaded = dl_manager.download(data_urls)
+                train_files.append(downloaded["train"])
+                test_files.append(downloaded["test"])
+            
+            return [
+                datasets.SplitGenerator(
+                    name=datasets.Split.TRAIN,
+                    gen_kwargs={
+                        "filepaths": train_files,
+                        "split": "train",
+                    },
+                ),
+                datasets.SplitGenerator(
+                    name=datasets.Split.TEST,
+                    gen_kwargs={
+                        "filepaths": test_files,
+                        "split": "test",
+                    },
+                ),
+            ]
         else:
-            # The data is stored in nested folders: data/{config_name}/train and data/{config_name}/test
-            # Use glob pattern to match any arrow files
-            data_files = {
-                "train": f"data/{self.config.name}/train/data-*-of-*.arrow",
-                "test": f"data/{self.config.name}/test/data-*-of-*.arrow",
+            # Single configuration
+            # Download the dataset_dict.json to understand the structure
+            dataset_dict_url = f"{self.config.data_dir}/dataset_dict.json"
+            dataset_dict_path = dl_manager.download(dataset_dict_url)
+            
+            # Download the actual data files
+            data_urls = {
+                "train": f"{self.config.data_dir}/train/data-00000-of-00001.arrow",
+                "test": f"{self.config.data_dir}/test/data-00000-of-00001.arrow",
             }
-        
-        # Download and resolve the files
-        downloaded_files = dl_manager.download_and_extract(data_files)
-        
-        return [
-            datasets.SplitGenerator(
-                name=datasets.Split.TRAIN,
-                gen_kwargs={
-                    "filepaths": downloaded_files["train"],
-                    "split": "train",
-                },
-            ),
-            datasets.SplitGenerator(
-                name=datasets.Split.TEST,
-                gen_kwargs={
-                    "filepaths": downloaded_files["test"],
-                    "split": "test",
-                },
-            ),
-        ]
+            
+            downloaded_files = dl_manager.download(data_urls)
+            
+            return [
+                datasets.SplitGenerator(
+                    name=datasets.Split.TRAIN,
+                    gen_kwargs={
+                        "filepaths": [downloaded_files["train"]],
+                        "split": "train",
+                    },
+                ),
+                datasets.SplitGenerator(
+                    name=datasets.Split.TEST,
+                    gen_kwargs={
+                        "filepaths": [downloaded_files["test"]],
+                        "split": "test",
+                    },
+                ),
+            ]
 
     def _generate_examples(self, filepaths, split):
-        # Flatten the list if it's a list of lists (happens with "all" config)
-        if isinstance(filepaths, list) and len(filepaths) > 0 and isinstance(filepaths[0], list):
-            # Flatten nested lists
-            flat_filepaths = []
-            for sublist in filepaths:
-                if isinstance(sublist, list):
-                    flat_filepaths.extend(sublist)
-                else:
-                    flat_filepaths.append(sublist)
-            filepaths = flat_filepaths
-        elif isinstance(filepaths, str):
+        # Load the Arrow files and yield examples
+        import pyarrow as pa
+        
+        # Handle both single filepath and list of filepaths
+        if isinstance(filepaths, str):
             filepaths = [filepaths]
         
-        # Generate examples from all arrow files
         example_id = 0
         for filepath in filepaths:
-            # Skip if filepath is None or empty
-            if not filepath:
-                continue
+            # Read the arrow file
+            with open(filepath, "rb") as f:
+                # Read arrow file using pyarrow
+                reader = pa.ipc.open_file(f)
+                table = reader.read_all()
+                
+                # Convert to pandas for easier iteration
+                df = table.to_pandas()
                 
-            # Load the Arrow file
-            try:
-                dataset = datasets.Dataset.from_file(filepath)
-                for row in dataset:
+                # Yield each row as an example
+                for idx, row in df.iterrows():
                     yield example_id, {
                         "input_sequence": row["input_sequence"],
                         "target": row["target"],
                     }
-                    example_id += 1
-            except Exception as e:
-                print(f"Warning: Could not load file {filepath}: {e}")
-                continue
+                    example_id += 1