Treat undefined correlation as zero correlation
This commit is contained in:
+1
-1
@@ -33,7 +33,7 @@ def calculate_correlation(df: pd.DataFrame, target_id: str | None = None) -> pd.
|
||||
def max_abs_corr(group: pd.DataFrame) -> pd.Series:
|
||||
corr_arr = np.abs(group.corr().to_numpy().copy())
|
||||
np.fill_diagonal(corr_arr, 0.0)
|
||||
return pd.Series(corr_arr.max(axis=0), index=group.columns)
|
||||
return pd.Series(corr_arr.max(axis=0), index=group.columns).fillna(0.0)
|
||||
|
||||
return df_bytes.groupby("Identifier")[available_cols].apply(max_abs_corr)
|
||||
|
||||
|
||||
Reference in New Issue
Block a user