rgwfuncs 0.0.13__py3-none-any.whl → 0.0.15__py3-none-any.whl

rgwfuncs/df_lib.py

@@ -1614,7 +1614,6 @@ def mask_against_dataframe_converse(df: pd.DataFrame, other_df: pd.DataFrame, co
 
     return df[~df[column_name].isin(other_df[column_name])]
 
-
 def union_join(df1: pd.DataFrame, df2: pd.DataFrame) -> pd.DataFrame:
     """
     Perform a union join, concatenating the two DataFrames and dropping duplicates.
@@ -1633,19 +1632,19 @@ def union_join(df1: pd.DataFrame, df2: pd.DataFrame) -> pd.DataFrame:
     # print("Initial df1 columns:", df1.columns)
     # print("Initial df2 columns:", df2.columns)
 
-    # Standardize columns by adding missing columns filled with NaN
+    # Standardize columns by adding missing columns filled with empty strings
     for col in df2.columns:
         if col not in df1:
-            df1[col] = pd.NA
+            df1[col] = ""
 
     for col in df1.columns:
         if col not in df2:
-            df2[col] = pd.NA
+            df2[col] = ""
 
     # print("Standardized df1 columns:", df1.columns)
     # print("Standardized df2 columns:", df2.columns)
 
-    # Check if columns match now
+    # Ensure they have the same columns after standardizing
     if set(df1.columns) != set(df2.columns):
         raise ValueError("Both DataFrames must have the same columns after standardizing columns")
 
@@ -1653,7 +1652,6 @@ def union_join(df1: pd.DataFrame, df2: pd.DataFrame) -> pd.DataFrame:
     result_df = pd.concat([df1, df2], ignore_index=True).drop_duplicates()
     return result_df
 
-
 def bag_union_join(df1: pd.DataFrame, df2: pd.DataFrame) -> pd.DataFrame:
     """
     Perform a bag union join, concatenating the two DataFrames without dropping duplicates.
@@ -1672,14 +1670,14 @@ def bag_union_join(df1: pd.DataFrame, df2: pd.DataFrame) -> pd.DataFrame:
     # print("Initial df1 columns:", df1.columns)
     # print("Initial df2 columns:", df2.columns)
 
-    # Standardize columns by adding missing columns filled with NaN
+    # Standardize columns by adding missing columns filled with empty strings
     for col in df2.columns:
         if col not in df1:
-            df1[col] = pd.NA
+            df1[col] = ""
 
     for col in df1.columns:
         if col not in df2:
-            df2[col] = pd.NA
+            df2[col] = ""
 
     # print("Standardized df1 columns:", df1.columns)
     # print("Standardized df2 columns:", df2.columns)

{rgwfuncs-0.0.13.dist-info → rgwfuncs-0.0.15.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: rgwfuncs
-Version: 0.0.13
+Version: 0.0.15
 Summary: A functional programming paradigm for mathematical modelling and data science
 Home-page: https://github.com/ryangerardwilson/rgwfunc
 Author: Ryan Gerard Wilson

rgwfuncs-0.0.15.dist-info/RECORD

@@ -0,0 +1,8 @@
+rgwfuncs/__init__.py,sha256=BP8Nh8ivyCCz8Ga-21JW3NWInJFOElKoIfRuioJRWbA,1076
+rgwfuncs/df_lib.py,sha256=VtCoaxx-ft864ChNZA3YEmvC8muVhFePqJX-j-zCzi0,64110
+rgwfuncs-0.0.15.dist-info/LICENSE,sha256=7EI8xVBu6h_7_JlVw-yPhhOZlpY9hP8wal7kHtqKT_E,1074
+rgwfuncs-0.0.15.dist-info/METADATA,sha256=ULUa8AcemlFS6tmNsG_Oxxvk2USzEsclQgznEC3QKNQ,32059
+rgwfuncs-0.0.15.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
+rgwfuncs-0.0.15.dist-info/entry_points.txt,sha256=j-c5IOPIQ0252EaOV6j6STio56sbXl2C4ym_fQ0lXx0,43
+rgwfuncs-0.0.15.dist-info/top_level.txt,sha256=aGuVIzWsKiV1f2gCb6mynx0zx5ma0B1EwPGFKVEMTi4,9
+rgwfuncs-0.0.15.dist-info/RECORD,,

rgwfuncs-0.0.13.dist-info/RECORD

@@ -1,8 +0,0 @@
-rgwfuncs/__init__.py,sha256=BP8Nh8ivyCCz8Ga-21JW3NWInJFOElKoIfRuioJRWbA,1076
-rgwfuncs/df_lib.py,sha256=2JU1Z0wjBrpcmLuJgY6-Bi1OFXHyNk1XeRGESkDI934,64077
-rgwfuncs-0.0.13.dist-info/LICENSE,sha256=7EI8xVBu6h_7_JlVw-yPhhOZlpY9hP8wal7kHtqKT_E,1074
-rgwfuncs-0.0.13.dist-info/METADATA,sha256=TZoSn8ANcHG2IOAW1SwNbu3ZnPKxdWbKzoCAXLTB06w,32059
-rgwfuncs-0.0.13.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
-rgwfuncs-0.0.13.dist-info/entry_points.txt,sha256=j-c5IOPIQ0252EaOV6j6STio56sbXl2C4ym_fQ0lXx0,43
-rgwfuncs-0.0.13.dist-info/top_level.txt,sha256=aGuVIzWsKiV1f2gCb6mynx0zx5ma0B1EwPGFKVEMTi4,9
-rgwfuncs-0.0.13.dist-info/RECORD,,