ValueError: On entry to SGESDD parameter number 12 had an illegal value

[ericustc]

python test_lfw.py

(71846, 5900) (71846, 1)
2015-10-27, 19:18:08 PCA training (n_components=2000)...
Traceback (most recent call last):
File "test_lfw.py", line 64, in
excute_train()
File "test_lfw.py", line 22, in excute_train
pca = PCA_Train(data, result_fold)
File "/opt/wjc/Joint-Bayesian/src/joint_bayesian.py", line 122, in PCA_Train
pca.fit(data)
File "/usr/lib64/python2.7/site-packages/sklearn/decomposition/pca.py", line 221, in fit
self._fit(X)
File "/usr/lib64/python2.7/site-packages/sklearn/decomposition/pca.py", line 271, in _fit
U, S, V = linalg.svd(X, full_matrices=False)
File "/usr/lib64/python2.7/site-packages/scipy/linalg/decomp_svd.py", line 104, in svd
full_matrices=full_matrices, overwrite_a=overwrite_a)
ValueError: On entry to SGESDD parameter number 12 had an illegal value

what's your method of solving the error?

[lufo816]

Use IncrementalPCA()

[HKUSTmji]

the same problem, even replacing the PCA method to the IncrementalPCA. Any suggestions?

[leandrosnm]

Hi, I had the same issue and solved it by using IncrementalPCA() and updating numpy to version 1.11.0 and scipy to version 0.17.0

It seems there's a bug somewhere in numpy and scipy older versions.

[zjz5250]

@ericustc, can you share the data? l can not download . thanks