Error using py.fooof.objs.fit.FOOOF/get_results (Matlab 2024a with Python 3.11)

[SvennoNito]

Hey guys, quickly wanted to report that with Matlab 2024a and Python 3.11.5 I get the following error. With Matlab 2023a and Python 3.8.2 it runs smoothly, so I can stay with the older Matlab version for now. Still thought it would be good to figure out what the problem is (never saw this kind of error).

Error using py.fooof.objs.fit.FOOOF/get_results
Same base class matlab.mixin.internal.indexing.Paren specified more than once for class py.fooof.data.data.FOOOFResults. Specify each base class
only once.

Error in fooof (line 59)
    fooof_results = fm.get_results();

Error in fooof_example_plot_model (line 25)
fooof_results = fooof(freqs, psd, f_range, settings, true);

[neuropil]

Getting the same error. Matlab 2024a and Python Version: "3.10"

[jiyunshin]

Same error with Matlab 2024 a and Python 3.11

[JohnnyVVParr]

I ran in to the same problem with MATLAB 2024a and Python 3.10/3.11.

Followed @SvennoNito suggestion and just regressed to MATLAB 2023a and Python 3.8.2 and it worked instantly following this guide: https://irenevigueguix.wordpress.com/2020/03/25/loading-python-into-matlab/

[dantingz]

Unfortunately` my linux server does not allow downgrading. Wrote a function to extract from the FOOOF without using get_results. Works for MATLAB 2024a and Python 3.11 now.

function fooof_results = fooof(freqs, power_spectrum, f_range, settings, return_model)

settings = fooof_check_settings(settings);

freqs_py = py.numpy.array(freqs);
power_py = py.numpy.array(power_spectrum);
f_range_py = py.list(f_range);


fm = py.fooof.FOOOF(settings.peak_width_limits, ...
                    settings.max_n_peaks, ...
                    settings.min_peak_height, ...
                    settings.peak_threshold, ...
                    settings.aperiodic_mode, ...
                    settings.verbose);


fm.fit(freqs_py, power_py, f_range_py);

% avoid fm.get_results()
fooof_results = struct();
fooof_results.aperiodic_params   = double(py.array.array('d', fm.aperiodic_params_));
fooof_results.peak_params = double(fm.peak_params_);
fooof_results.gaussian_params    = double(fm.gaussian_params_);
fooof_results.r_squared          = double(fm.r_squared_);
fooof_results.error              = double(fm.error_);
fooof_results.freq_range         = f_range;

if exist('return_model', 'var') && return_model
    
    fooof_results.freqs            = double(py.array.array('d', fm.freqs));
    fooof_results.power_spectrum   = double(py.array.array('d', fm.power_spectrum));


    fooof_results.fooofed_spectrum = double(py.array.array('d', fm.fooofed_spectrum_));
    fooof_results.ap_fit = double(py.array.array('d', py.getattr(fm, '_ap_fit')));

    % Recalculate r_squared manually
    coefs = corrcoef(fooof_results.power_spectrum, fooof_results.fooofed_spectrum);
    if numel(coefs) == 4
        fooof_results.r_squared = coefs(2);
    else
        fooof_results.r_squared = NaN;
    end
end

end