Predictive Power Score PPS in Matlab?

Is there a Matlab implementation of the Predictive Power Score PPS, equivalent to the one available in Python (https://towardsdatascience.com/rip-correlation-introducing-the-predictive-power-score-3d90808b9598)?

Answer 1

[Ayush Aniket]

As of the R2024b release, MATLAB does not include a function for calculating the Predictive Power Score (PPS). However, you can utilize the MATLAB Engine API for Python to call the ppscore library from Python. Below is an example code:

% Sample data as a MATLAB table
data = table([1; 2; 3; 4; 5], [10; 20; 30; 40; 50], [0; 1; 0; 1; 0], ...
             'VariableNames', {'Feature1', 'Feature2', 'Target'});

% Convert MATLAB table to Python DataFrame
pyDataFrame = py.pandas.DataFrame(data);

% Import the ppscore library
ppscore = py.importlib.import_module('ppscore');

% Calculate the PPS for a specific feature and target
pps_score = ppscore.score(pyDataFrame, 'Feature1', 'Target');

Note that for implementing this you need to ensure that you have Python installed on your system with the ppscore and pandas libraries available.

Refer to the following documentation link to read about how to use a pandas DataFrame in MATLAB.:

https://www.mathworks.com/help/matlab/matlab_external/python-pandas-dataframes.html