make_watershed_bem() does not make a BEM usable by setup_volume_source_space()

✉️ Mailing List Archive (read-only)
1

External Email - Use Caution

Hi,

At line 1182 of
https://github.com/mne-tools/mne-python/blob/maint/0.19/mne/bem.py
mne.bem.make_watershed_bem() is creating the BEM FIF file using
FIFF.FIFFV_BEM_SURF_ID_HEAD (==4). When using this BEM file for source
modeling mne.setup_volume_source_space() expect
FIFF.FIFFV_BEM_SURF_ID_BRAIN (==1) (line 1705
https://github.com/mne-tools/mne-python/blob/master/mne/source_space.py)
and then crash because this surface id is not found in the BEM (stack trace
below).

I used

mne.bem.make_watershed_bem(fs_subject, subjects_dir=subjects_dir,
                           overwrite=False, show=True)

to make the BEM and

vol_src = setup_volume_source_space(
    subject, mri=fname_aseg, pos=10.0, bem=fname_model,
    add_interpolator=False, volume_label=labels_vol,
subjects_dir=subjects_dir)

to build the source space. Do you think it is a bug (I'll make a ticker if
so), or does it look like I used these functions not correctly?

Best,

Christian

2

External Email - Use Caution

That function just makes the surfaces, plus the subject's head FIF. I
suspect you've tried to use `subject-head.fif` as your BEM, which will not
work.

You need to use `mne.make_bem_model` and `mne.make_bem_solution` to
actually make the BEM itself from the surfaces `make_watershed_bem`
produced, choosing along the way if you want a 1-layer or 3-layer model.
You can then save these to disk and load them when you need them.

Eric

3

External Email - Use Caution

Hi,
   Seasons' greetings.
           I am new to MNE and also to python. How to install mne-features
in anaconda navigator. I tried with pip install mne-features, but shows the
following error:

(base) C:\Users\Admin>pip install mne-features
Collecting mne-features
  Could not find a version that satisfies the requirement mne-features
(from versions: )
No matching distribution found for mne-features

the following Requirement are also satisfied:

These are the dependencies to use MNE-Features:

   - numpy (>=1.8)
   - matplotlib (>=1.3)
   - scipy (>=0.19)
   - numba (>=0.37)
   - scikit-learn (>=0.19)
   - mne (>=0.14)
   - PyWavelets (>=0.5.2)
   - pandas (>=0.20)

please help me out.
regards,

*M.Bhuvaneshwari,*
*Research Scholar,*
*Karunya Institute of technology and Sciences*

*Coimbatore.*

4

External Email - Use Caution

hi,

mne-features has never been pushed to pypi. You should do:

pip install git+
https://github.com/mne-tools/mne-features.git#egg=mne_features

to install from the master branch.

Readme of mne-features needs to be updated or package should be released and
pushed to pypi.

Alex