scikit-learn version

Hi all,

I am running the classify-sklearn commond, but I got this error:

(qiime2) yiggrite@MWMJ07RHBF:~/qiime2/muxed-se-barcode-in-seq$ qiime feature-classifier classify-sklearn
13-8-99-> --i-classifier gg-13-8-99-515-806-nb-classifier.qza \

--i-reads rep-seqs.qza
--o-classification taxonomy.qza
etadata tabulate
--m-input-file taxonomy.qza
--o-visualization taxonomy.qzv
Plugin error from feature-classifier:

The scikit-learn version (0.24.1) used to generate this artifact does not match the current version of scikit-learn installed (0.23.1). Please retrain your classifier for your current deployment to prevent data-corruption errors.

Debug info has been saved to /tmp/qiime2-q2cli-err-s60qysis.log

I tried to update qiime2, i met this problem:

(qiime2) yiggrite@MWMJ07RHBF:~/qiime2/muxed-se-barcode-in-seq$ conda update conda
Retrieving notices: ...working... done

PackageNotInstalledError: Package is not installed in prefix.
prefix: /home/yiggrite/miniconda3/envs/qiime2
package name: conda

(qiime2) yiggrite@MWMJ07RHBF:~/qiime2/muxed-se-barcode-in-seq$ cd ..
(qiime2) yiggrite@MWMJ07RHBF:~/qiime2$ conda update conda

PackageNotInstalledError: Package is not installed in prefix.
prefix: /home/yiggrite/miniconda3/envs/qiime2
package name: conda

(qiime2) yiggrite@MWMJ07RHBF:~/qiime2$ cd ..
(qiime2) yiggrite@MWMJ07RHBF:~ cd miniconda3 (qiime2) yiggrite@MWMJ07RHBF:~/miniconda3 cd envs
(qiime2) yiggrite@MWMJ07RHBF:~/miniconda3/envs$ cd qiime2
(qiime2) yiggrite@MWMJ07RHBF:~/miniconda3/envs/qiime2$ conda update conda

PackageNotInstalledError: Package is not installed in prefix.
prefix: /home/yiggrite/miniconda3/envs/qiime2
package name: conda

(qiime2) yiggrite@MWMJ07RHBF:~/miniconda3/envs/qiime2$

How can I solve this problem?

Thanks

Hi @Yiggrite,

It looks like you are attempting to update conda from within one of your activated conda environments. You'll want to first run conda deactivate and make sure that you are in your base environment before running conda update conda.

From there, you should be able to either update your current qiime2 environment with the correct scikit-learn version, or create a different qiime2 environment with that version.

Cheers :lizard:

1 Like

Hi,

I have updated my qiime2, but the scikit-learn version is still 23.1. How can I update the scikit-learn?

Thanks

(qiime2) yiggrite@MWMJ07RHBF:~/qiime2/muxed-se-barcode-in-seq$ conda deactivate
(base) yiggrite@MWMJ07RHBF:~/qiime2/muxed-se-barcode-in-seq$ conda update conda
Retrieving notices: ...working... done
Channels:

  • defaults
    Platform: linux-64
    Collecting package metadata (repodata.json): done
    Solving environment: done

All requested packages already installed.

(base) yiggrite@MWMJ07RHBF:~/qiime2/muxed-se-barcode-in-seq$ conda activate qiime2
(qiime2) yiggrite@MWMJ07RHBF:~/qiime2/muxed-se-barcode-in-seq$ conda update qiime2
Channels:

  • defaults
  • conda-forge
  • bioconda
  • qiime2/label/r2020.8
    Platform: linux-64
    Collecting package metadata (repodata.json): done
    Solving environment: done

All requested packages already installed.

(qiime2) yiggrite@MWMJ07RHBF:~/qiime2/muxed-se-barcode-in-seq$ conda list scikit-learn

packages in environment at /home/yiggrite/miniconda3/envs/qiime2:

Name Version Build Channel

scikit-learn 0.23.1 py36h7ea95a0_0
(qiime2) yiggrite@MWMJ07RHBF:~/qiime2/muxed-se-barcode-in-seq$

Hello @Yiggrite. It is more or less impossible to update QIIME 2 or any of its dependencies to new versions "in place" without creating a whole new conda environment. This is because QIIME 2 has a large number of 3rd party dependencies and requires specific versions of many of those dependencies and conda figures out which versions are compatible with each other and creates an environment file with those versions.

In order to update scikit learn to 0.24.1, you are going to want to install QIIME 2 2023.9. In order to install QIIME 2 2023.9, you should create a new conda environment following the instructions found here. You will want the "Amplicon" distribution.

2 Likes

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.