remove uneeded code and dependencies

Author: david-cortes-intel

envs/conda-env-rapids.yml

@@ -10,8 +10,7 @@ dependencies:
   # sklbench dependencies
   - scikit-learn
   - pandas
-  - tabulate
-  - fastparquet
+  - pyarrow
   - h5py
   - openpyxl
   - tqdm

envs/conda-env-sklearn.yml

@@ -6,14 +6,11 @@ dependencies:
   - xgboost
   - catboost
   - lightgbm
-  - faiss-cpu
-  - modin-all
   - scikit-learn-intelex
   # sklbench dependencies
   - scikit-learn
   - pandas
-  - tabulate
-  - fastparquet
+  - pyarrow
   - h5py
   - openpyxl
   - tqdm

envs/requirements-sklearn.txt

@@ -2,16 +2,13 @@
 xgboost
 catboost
 lightgbm
-faiss-cpu
-modin[all]
 scikit-learn-intelex
 dpctl
 dpnp
 # sklbench dependencies
 scikit-learn
 pandas
-tabulate
-fastparquet
+pyarrow
 h5py
 openpyxl
 tqdm

sklbench/datasets/common.py

@@ -101,7 +101,7 @@ def save_data_to_cache(data: Dict, data_cache: str, data_name: str):
                 for column in list(data_compoment.columns)
             ]
             data_compoment.to_parquet(
-                component_filepath, engine="fastparquet", compression="snappy"
+                component_filepath
             )
         elif isinstance(data_compoment, csr_matrix):
             component_filepath += ".csr.npz"

sklbench/datasets/transformer.py

@@ -57,24 +57,6 @@ def convert_data(data, dformat: str, order: str, dtype: str, device: str = None)
         import dpctl.tensor
 
         return dpctl.tensor.asarray(data, dtype=dtype, order=order, device=device)
-    elif dformat.startswith("modin"):
-        if dformat.endswith("ray"):
-            os.environ["MODIN_ENGINE"] = "ray"
-        elif dformat.endswith("dask"):
-            os.environ["MODIN_ENGINE"] = "dask"
-        elif dformat.endswith("unidist"):
-            os.environ["MODIN_ENGINE"] = "unidist"
-            os.environ["UNIDIST_BACKEND"] = "mpi"
-        else:
-            logger.info(
-                "Modin engine is unknown or not specified. Default engine will be used."
-            )
-
-        import modin.pandas as modin_pd
-
-        if data.ndim == 1:
-            return modin_pd.Series(data)
-        return modin_pd.DataFrame(data)
     elif dformat == "cudf":
         import cudf