Add train test split, benchmark split

airbnb_multicity.py

@@ -4,26 +4,43 @@ import pyarrow.parquet as pq
 
 
 DESCRIPTION = "The dataset contains Airbnb data from 80 capitals and major cities all around the world."
-DATA_URL="https://huggingface.co/datasets/kraina/airbnb_multicity/resolve/main/data/all_airbnb.parquet"
+# DATA_URL="https://huggingface.co/datasets/kraina/airbnb_multicity/resolve/main/data/all_airbnb.parquet"
+
+DATA_DIRS = ["benchmark", "all"]
+
 
 class AirbnbDatasetConfig(datasets.BuilderConfig):
     """BuilderConfig """
 
-    def __init__(self, **kwargs):
+    def __init__(self, data_url, **kwargs):
         """BuilderConfig.
         Args:
           **kwargs: keyword arguments forwarded to super.
         """
         super(AirbnbDatasetConfig, self).__init__(**kwargs)
+        self.data_url = data_url
         
 
 class AirbnbDataset(datasets.ArrowBasedBuilder):
     BUILDER_CONFIG_CLASS = AirbnbDatasetConfig
+    DEFAULT_CONFIG_NAME = "benchmark"
+
+    BUILDER_CONFIGS = [
+        AirbnbDatasetConfig(
+            name = dir_name,
+            description = "This is the official train test split for Airbnb Datatset in h3 resolution = 8. Benchmark cities are: Paris, London, Rome, Melbourne, New York City, Amsterdam."+DESCRIPTION,
+            data_url={
+                "train": f"https://huggingface.co/datasets/kraina/airbnb_multicity/resolve/main/data/{dir_name}/airbnb_train.parquet",
+                "test": f"https://huggingface.co/datasets/kraina/airbnb_multicity/resolve/main/data/{dir_name}/airbnb_test.parquet"
+            }
+        )
+        for dir_name in DATA_DIRS
+    ]
 
     def _info(self):
         return datasets.DatasetInfo(
             # This is the description that will appear on the datasets page.
-            description=DESCRIPTION,
+            description=self.config.description,
             homepage="https://insideairbnb.com/",
             citation="",
             # This defines the different columns of the dataset and their types
@@ -53,15 +70,18 @@ class AirbnbDataset(datasets.ArrowBasedBuilder):
 
 
     def _split_generators(self, dl_manager: datasets.download.DownloadManager):
-        downloaded_files = dl_manager.download(DATA_URL)
-        return [
-            datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={'filepath': downloaded_files})
-        ]
+        downloaded_files = dl_manager.download(self.config.data_url)
 
+        return [
+                datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={'filepath': downloaded_files["train"]}),
+                datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={'filepath': downloaded_files["test"]})
+            ]
     # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
     def _generate_tables(self, filepath):
         with open(filepath, mode="rb") as f:
             parquet_file = pq.ParquetFile(source=filepath)
             for batch_idx, record_batch in enumerate(parquet_file.iter_batches()):
-                pa_table = pa.Table.from_batches([record_batch])
+                df = record_batch.to_pandas()
+                df.reset_index(drop=True, inplace=True)
+                pa_table = pa.Table.from_pandas(df)
                 yield f"{batch_idx}", pa_table

data/all/airbnb_test.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:321df503893d2d07d49a0d776f1cfe1b0cfc6df664d6909bee89bad4972f54d2
+size 49758512

data/all/airbnb_train.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41f9754acce3d42c509294ccddca707be57dc74d5e3eeda99974012c8c649e4d
+size 199355005

data/benchmark/airbnb_test.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62e71598833914a3a19967b49d89d1995fdd4cb93e51744ea9fabcf2538eacda
+size 10729670

data/benchmark/airbnb_train.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61c480b8839a307c711eaebe6b4c612cf7fb57376fddee0d714795664f2bf8dd
+size 39544306