Merge pull request #20 from allegro/update-dependencies

Remove pandas and update dependencies

allms/utils/io_utils.py

@@ -1,35 +1,44 @@
+import csv
 import logging
 from pathlib import Path
-from typing import List, Optional, Union
+from typing import Any, Dict, List, Optional, Union, OrderedDict
 
 import fsspec
-import pandas as pd
 
 from allms.constants.input_data import IODataConstants
 from allms.domain.input_data import InputData
 
 logger = logging.getLogger(__name__)
 
 
-def load_data(
+def load_csv(
         path: str,
         limit: Optional[int] = None
-) -> List[InputData]:
+) -> List[OrderedDict[Any, Any]]:
     logger.info(f"Loading test data from {path}")
-    input_df = pd.read_csv(path)
-    input_df = input_df.head(limit) if limit else input_df
-    return load_input_data(input_df)
+    with open(path, mode='r') as csv_file:
+        csv_reader = csv.DictReader(csv_file)
+        data = list(csv_reader)
+        return data[:limit] if limit else data
 
 
-def load_input_data(input_df: pd.DataFrame) -> List[InputData]:
+def load_csv_to_input_data(path: str, limit: Optional[int] = None) -> List[InputData]:
+    csv_data = load_csv(path, limit=limit)
     return list(
         map(
-            lambda row: InputData(input_mappings=row[1].drop(IODataConstants.ID).to_dict(), id=str(row[1].id)),
-            input_df.iterrows()
+            lambda row: InputData(input_mappings=drop_dict_key(row, IODataConstants.ID),
+                                  id=str(row[IODataConstants.ID])),
+            csv_data
         )
     )
 
 
+def drop_dict_key(dictionary: Dict[Any, Any], key: Any) -> Dict[Any, Any]:
+    dict_copy = dictionary.copy()
+    dict_copy.pop(key)
+    return dict_copy
+
+
 def load_credentials(path: Union[str, Path]) -> str:
     with fsspec.open(path, "r") as credentials_file:
         return credentials_file.readline()