qolmat.benchmark.metrics.mean_difference_correlation_matrix_numerical_features

qolmat.benchmark.metrics.mean_difference_correlation_matrix_numerical_features(df1: DataFrame, df2: DataFrame, df_mask: DataFrame, use_p_value: bool = True) Series[source]

Compute the mean absolute of differences.

Computed between the correlation matrices of df1 and df2. based on Pearson correlation coefficient or p-value for testing non-correlation.

Parameters
df1pd.DataFrame

true dataframe

df2pd.DataFrame

predicted dataframe

df_maskpd.DataFrame

Elements of the dataframes to compute on

use_p_valuebool, optional

use the p-value instead of the correlation coefficient, by default True

Returns
pd.Series

Mean absolute of differences for each feature