MNE 1.0 in conda-forge?

💬 Support & Discussions
1

Hi MNE team. First of all, thanks a lot for your work and congrats to the 1.0 version!

Since I am an anaconda user, my question is whether you plan to update MNE in the conda-forge channel as well?

2

Hello @stefan, what exactly do you mean? MNE-Python 1.0 has been on conda-forge since the release, and our new installers are in fact based on these packages. See:

https://anaconda.org/conda-forge/mne

3

Hi Richard, thanks for the quick response. I should have double-checked. My environment did not use the “main” label by default, so I ended up with the linux-64 0.23.4 version.

1 Like
4

Hi Stefan,

We are currently facing the same problem. Could you tell me what exactly you changed to make the installation work?

Thanks,
Laura

5

Hi Laura,

I used

conda install -c conda-forge/label/main mne

to install the v1.0.2 noarch version in a newly created testing environment (still have to fix my default environment though…). Without explicitly stating the main label the installer used the linux-64 version (os is Ubuntu 20.04).

1 Like
6

Very interesting. Could you please share the output of conda info?

7

Sure.

In a newly created environment, mne installed using

conda install -c conda-forge mne

conda info:

     active environment : test_a
    active env location : /home/plkn/miniconda3/envs/test_a
            shell level : 3
       user config file : /home/plkn/.condarc
 populated config files : 
          conda version : 4.12.0
    conda-build version : not installed
         python version : 3.8.13.final.0
       virtual packages : __cuda=11.4=0
                          __linux=5.13.0=0
                          __glibc=2.31=0
                          __unix=0=0
                          __archspec=1=x86_64
       base environment : /home/plkn/miniconda3  (writable)
      conda av data dir : /home/plkn/miniconda3/etc/conda
  conda av metadata url : None
           channel URLs : https://repo.anaconda.com/pkgs/main/linux-64
                          https://repo.anaconda.com/pkgs/main/noarch
                          https://repo.anaconda.com/pkgs/r/linux-64
                          https://repo.anaconda.com/pkgs/r/noarch
          package cache : /home/plkn/miniconda3/pkgs
                          /home/plkn/.conda/pkgs
       envs directories : /home/plkn/miniconda3/envs
                          /home/plkn/.conda/envs
               platform : linux-64
             user-agent : conda/4.12.0 requests/2.27.1 CPython/3.8.13 Linux/5.13.0-40-generic ubuntu/20.04.4 glibc/2.31
                UID:GID : 1000:1000
             netrc file : None
           offline mode : False

conda list mne:

# packages in environment at /home/plkn/miniconda3/envs/test_a:
#
# Name                    Version                   Build  Channel
mne                       0.23.4               ha770c72_1    conda-forge
mne-base                  0.23.4           py39ha770c72_1    conda-forge
8

I also tried

conda config --set channel_priority false

which did not change the package selection.

Explicitly stating the version number works as well:

conda install -c conda-forge mne=1.0.2
9

Thank you, @stefan, I’ll look into this next week so that we’ll hopefully get to the bottom of this. Thank you for sharing this useful information!

10

We’re now tracking this on GitHub:

I may get back to you and ask for further info :slight_smile:

11

@stefan

Could you please share with me the output of

conda config --show channel_priority

on the system that shows the problem?

Thanks!

12

channel_priority is “flexible”. In my standard environment as well as in newly created ones.

13

Thank you, @stefan! This default value is what’s causing the problem. We will amend the installation instructions to work around this.

cc @marsipu

14

We’re working on updating the docs: