search

jmschrei / apricot

apricot implements submodular optimization for the purpose of selecting subsets of massive data sets to train machine learning models quickly. See the documentation page: https://apricot-select.readthedocs.io/en/latest/index.html
MIT License
499 stars 48 forks source link

bidirectional optimization with Graph Cut ValueError #35

Closed M-A-Hassan closed 2 years ago

M-A-Hassan commented 2 years ago

Thanks for the amazing package and documentation: I am facing issue when I try to use bidirectional optimization with graph cut GraphCutSelection(N_Samples_Per_Class, metric='euclidean', optimizer='bidirectional').fit_transform(.......... ValueError: zero-size array to reduction operation maximum which has no identity

Note that using other optimizers worked just fine GraphCutSelection(N_Samples_Per_Class, metric='euclidean', optimizer='lazy').fit_transform(.......... GraphCutSelection(N_Samples_Per_Class, metric='euclidean', optimizer='two-stage').fit_transform(..........

Your help is appreciated

jmschrei commented 2 years ago

Hi @M-A-Hassan. It looks like I never finished the Bidirectional optimizer and I should remove it. Sorry for the inconvenience.

M-A-Hassan commented 2 years ago

@jmschrei Noted. Thanks