Closed ikeuchi-screen closed 12 months ago
factor_analyzer
used in lina.py
is using scipy.minimize
.
lingam/lina.py:316: in fit
cfa.fit(X, G_sign)
/usr/local/lib/python3.10/site-packages/factor_analyzer/confirmatory_factor_analyzer.py:742: in fit
res = minimize(
@YanaZeng I don't understand the LiNA algorithm very well. As mentioned above, it may be a problem with the factor_analyzer library. Do you know how we can solve this problem?
Ikeuchi-san, I try to run the code, but without this error. Maybe it's due to the versions of scipy or factor_analyzer.
I used factor_analyzer 0.4.1 and scipy 1.9.3. And my 'x0' here is a 1d vector. Could you print your 'x0'?
Hi, @YanaZeng Thank you for your confirmation! I believe this is caused by a change in scipy-1.11.0, as mentioned in my opening post. (factor_analyzer-0.4.1, scipy-1.11.0)
The test code execution resulted in the following x0
values.
This shape is a 2d-array.
x0 = array([[1. ],
[1. ],
[0. ],
[0. ],
[0. ],
[0. ],
[0. ],
[0. ]... [0.05],
[0.05],
[0.05],
[0.05],
[0.05],
[0.05],
[0.05],
[0.05]])
The results of the test code execution can be found at the following link: https://github.com/cdt15/lingam/actions/runs/5376105848/jobs/9752810059#step:6:155
@YanaZeng
I have made a pull request to factor_analyzer
GitHub on how to fix this problem.
https://github.com/EducationalTestingService/factor_analyzer/pull/124
However, I don't think it will be merged for a while, so I will fix it so that the LiNA test will not run.
Get it, Ikeuchi-san! お疲れ様でした.
I close this issue because the PullRequest for factor_analyzer has been merged. https://github.com/EducationalTestingService/factor_analyzer/pull/124
I get the following error when I run the lina.py test.
The following changes in scipy-1.11.0 may have affected this.
SciPy 1.11.0 Release Notes