ikarmann

Different results with Tetrad and causal-learn implementations.

6

Dear developers, I have been trying to use causal learn models with fMRI data. I used both the Tetrad implementation (using pyTetrad and JPype) and the causal-learn implementations. What surprises...

The in the eBIC calculation (function `ebic` in metrics.py), l_theta is multiplied by `n_features / 2.` whereas the log-likelihood calculation implies multiplying by `n_samples / 2` as in the original...

### What happened? Processing (multiband) data from ADNI3 I noticed many subjects showed failure to coregister the fmap reference (magnitude image) with de EPI image, resulting in a very problematic...