supplementary / illustrative variables implemented?

[mluerig]

I was wondering whether supplementary variables are implemented (corresponding to num.group.sup argument in factominer)?

Also: I really appreciate your efforts with this - great initiative, great package, great name!

[MaxHalford]

Hey there @mluerig! I appreciate the kind words.

Yes, Prince supports supplementary variables:

import prince

dataset = prince.datasets.load_energy_mix(year=2019, normalize=True)
pca = prince.PCA()
pca.fit(dataset, supplementary_columns=['hydro', 'wind', 'solar'])
pca.column_correlations

This is unit tested against FactoMineR, and should provide exactly the same results 👌

[mluerig]

shoot, forgot to say that I meant MFA - doesn't seem mfa.fit accepts supplementary_columns?

[MaxHalford]

Ah indeed, Prince does not yet support supplementary variables for MFA :/

[mluerig]

too bad. probably not high on your priority list, but looking at the code for PCA and seeing how MFA is implemented it might not be too hard to implement supporting variables there as well?

let me know if you have any plans to do so, otherwise I could give it a shot sometime next year...

[MaxHalford]

Indeed I don't think it should be too difficult. I will try to take a look over the Xmas holidays!

[MaxHalford]

Progress in #185!

[MaxHalford]

@mluerig I just released v0.15. I completely revamped MFA: it now properly inherits from PCA, and thus has all the latter's methods. Supplementary rows and supplementary groups are now supported.

Please let me know if you're looking for anything particular! I can strike the iron while it's hot.

Also, if I may ask, what are you using MFA for?