decompy.rankmethods package

Submodules

decompy.rankmethods.bayes module

decompy.rankmethods.bayes.rank_hoffbayes(Y, svdfunc, gibbstype='fixed', kmax=20, verbose=False)

Rank estimation using Bayesian method - Reference: https://www.jstor.org/stable/27639896

decompy.rankmethods.cvrank module

decompy.rankmethods.cvrank.rank_bcv(X: ndarray, svdfunc, ks=array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), cvs=(None, None), verbose=False, metric='MSE')

” Rank estimation using Bi-cross validation technique Reference - https://doi.org/10.1214/08-AOAS227

decompy.rankmethods.cvrank.rank_gabriel(X: ndarray, svdfunc, ks=array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), cvs=100, verbose=False, metric='MSE')

Rank estimation using Gabriel style cross validation - http://www.numdam.org/item/JSFS_2002__143_3-4_5_0/

decompy.rankmethods.cvrank.rank_separate_rowcol(X: ndarray, svdfunc, ks=array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), cvs=100, verbose=False, metric='MSE')

Rank estimation using separate Row and Column deletion - https://www.jstor.org/stable/1267581

decompy.rankmethods.penalized module

decompy.rankmethods.penalized.rank_DIC(X: ndarray, svdfunc, alpha=0.5, ks=array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), verbose=False)

Rank estimation using divergence information criterion Reference:

decompy.rankmethods.penalized.rank_bai_ng(X: ndarray, svdfunc, ks=array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), criterion='PC1', verbose=False)

Rank estimation using information criterion by Bai and Ng. Reference - https://doi.org/10.1111/1468-0262.00273

decompy.rankmethods.penalized.rank_classical_ic(X: ndarray, svdfunc, ks=array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), criterion='AIC', verbose=False)

Rank estimation using classical criterion: AIC - Akaike’s Information Criteria (https://link.springer.com/chapter/10.1007/978-1-4612-1694-0_15) BIC - Bayesian Information Criteria

decompy.rankmethods.penalized.rank_elbow(X: ndarray, svdfunc, maxrank=30)

Rank estimation using Elbow Method

Module contents

decompy.rankmethods.rank_DIC(X: ndarray, svdfunc, alpha=0.5, ks=array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), verbose=False)

Rank estimation using divergence information criterion Reference:

decompy.rankmethods.rank_bai_ng(X: ndarray, svdfunc, ks=array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), criterion='PC1', verbose=False)

Rank estimation using information criterion by Bai and Ng. Reference - https://doi.org/10.1111/1468-0262.00273

decompy.rankmethods.rank_bcv(X: ndarray, svdfunc, ks=array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), cvs=(None, None), verbose=False, metric='MSE')

” Rank estimation using Bi-cross validation technique Reference - https://doi.org/10.1214/08-AOAS227

decompy.rankmethods.rank_classical_ic(X: ndarray, svdfunc, ks=array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), criterion='AIC', verbose=False)

Rank estimation using classical criterion: AIC - Akaike’s Information Criteria (https://link.springer.com/chapter/10.1007/978-1-4612-1694-0_15) BIC - Bayesian Information Criteria

decompy.rankmethods.rank_elbow(X: ndarray, svdfunc, maxrank=30)

Rank estimation using Elbow Method

decompy.rankmethods.rank_gabriel(X: ndarray, svdfunc, ks=array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), cvs=100, verbose=False, metric='MSE')

Rank estimation using Gabriel style cross validation - http://www.numdam.org/item/JSFS_2002__143_3-4_5_0/

decompy.rankmethods.rank_hoffbayes(Y, svdfunc, gibbstype='fixed', kmax=20, verbose=False)

Rank estimation using Bayesian method - Reference: https://www.jstor.org/stable/27639896

decompy.rankmethods.rank_separate_rowcol(X: ndarray, svdfunc, ks=array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]), cvs=100, verbose=False, metric='MSE')

Rank estimation using separate Row and Column deletion - https://www.jstor.org/stable/1267581