mapca error

[rhl508]

Hi

I'm using tedana version 23.0.2 and it gets quite far through and then errors at the pca. It looks like it's an issue with mapca. I'm using version 0.0.4 but I tried downgrading to 0.0.3 and that didn't work.

INFO     io:__init__:154 Generating figures directory: /scratch/groups/Projects/P1465/data/sub-002/ses-1/derivatives/tedana/prf/run-1/figures
INFO     tedana:tedana_workflow:564 Using user-defined mask
INFO     tedana:tedana_workflow:612 Computing T2* map
INFO     combine:make_optcom:191 Optimally combining data with voxel-wise T2* estimates
INFO     tedana:tedana_workflow:637 Writing optimally combined data set: /scratch/groups/Projects/P1465/data/sub-002/ses-1/derivatives/tedana/prf/run-1/desc-optcom_bold.nii.gz
INFO     pca:tedpca:202 Computing PCA of optimally combined multi-echo data with selection criteria: aic

Traceback (most recent call last):
  File "/home/r/rhl508/venvs/tedana/bin/tedana", line 8, in <module>
    sys.exit(_main())
  File "/home/r/rhl508/venvs/tedana/lib/python3.9/site-packages/tedana/workflows/tedana.py", line 918, in _main
    tedana_workflow(**kwargs, tedana_command=tedana_command)
  File "/home/r/rhl508/venvs/tedana/lib/python3.9/site-packages/tedana/workflows/tedana.py", line 641, in tedana_workflow
    dd, n_components = decomposition.tedpca(
  File "/home/r/rhl508/venvs/tedana/lib/python3.9/site-packages/tedana/decomposition/pca.py", line 214, in tedpca
    _ = ma_pca.fit_transform(data_img, mask_img)
  File "/home/r/rhl508/venvs/tedana/lib/python3.9/site-packages/mapca/mapca.py", line 478, in fit_transform
    self._fit(img, mask, subsample_depth=subsample_depth)
  File "/home/r/rhl508/venvs/tedana/lib/python3.9/site-packages/mapca/mapca.py", line 404, in _fit
    self.n_features_ = ppca.n_features_
AttributeError: 'PCA' object has no attribute 'n_features_'

Any advice appreciated. Thanks!

[rhl508]

Received an answer from a colleague that worked. It looks like 'nfeatures' was changed to 'n_featuresin' in newer versions of scikit-learn, so they changed mapca to match this.

I downgraded scikit-learn to 1.1.2 and that fixed it

[eurunuela]

Thank you for reporting this.

First of all, glad you found a workaround!

And second, we may want to open an issue to get this fixed for newer versions of scikit learn.

[eurunuela]

Ok, my bad. It's been a few crazy months. I already opened a pull request in maPCA that fixes this but haven't finished making other changes to merge it. Consider it done next week.

[rhl508]

Thanks so much!

On Fri 26 Jan 2024, 10:18 Eneko Uruñuela, @.***> wrote:

Ok, my bad. It's been a few crazy months. I already opened a pull request in maPCA https://github.com/ME-ICA/mapca/pull/59 that fixes this but haven't finished making other changes to merge it. Consider it done next week.

— Reply to this email directly, view it on GitHub https://github.com/ME-ICA/tedana/issues/1019#issuecomment-1911804391, or unsubscribe https://github.com/notifications/unsubscribe-auth/BFTGOQBP2O6OTXXIP6QOUILYQN7GTAVCNFSM6AAAAABCKNQRFKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMJRHAYDIMZZGE . You are receiving this because you modified the open/close state.Message ID: @.***>

[bedardp2]

Hi I still have the error as above. Is this normal?
I'm using tedana version 23.0.2

Also, (or) how can I downgrade to scikit-learn to 1.1.2 as [rhl508] mentioned previously? I'm using tedana within AFNI (afni_proc) and I had error, so I'm trying at CLI (tedana -d pb03.890012.r01.e01.volreg+tlrc.HEAD pb03.890012.r01.e02.volreg+tlrc.HEAD pb03.890012.r01.e03.volreg+tlrc.HEAD -e 19 38.21 57.42 --mask mask_group+tlrc.HEAD --out-dir tedana_r01 --convention orig --overwrite)

I don't see how I can change this I'm really not an expert with python!

I've attached the error message I received error.txt

thank you patrick

[eurunuela]

Hi Patrick, thank you for reporting this.

Could you try updating maPCA to the latest version? Version 0.0.5 fixed this issue.