feature-classifier-issue

I am trying to use silva classifier on the server, I have tried different classifiers but it didn't solve this issues

I always got
Plugin error from feature-classifier:

The scikit-learn version (0.20.2) used to generate this artifact does not match the current version of scikit-learn installed (1.4.2). Please retrain your classifier for your current deployment to prevent data-corruption errors.

Debug info has been saved to /tmp/qiime2-q2cli-err-iap__i4m.log

I have qiime 2024.5 version

qiime info

System versions

Python version: 3.9.19

QIIME 2 release: 2024.5

QIIME 2 version: 2024.5.1

q2cli version: 2024.5.0

Installed plugins

alignment: 2024.5.0

composition: 2024.5.0

cutadapt: 2024.5.0

dada2: 2024.5.0

deblur: 2024.5.0

demux: 2024.5.0

diversity: 2024.5.1

diversity-lib: 2024.5.0

emperor: 2024.5.0

feature-classifier: 2024.5.0

feature-table: 2024.5.0

fragment-insertion: 2024.5.0

longitudinal: 2024.5.0

metadata: 2024.5.0

phylogeny: 2024.5.0

quality-control: 2024.5.0

quality-filter: 2024.5.0

rescript: 2024.5.1

sample-classifier: 2024.5.0

taxa: 2024.5.0

types: 2024.5.0

vsearch: 2024.5.0

I have found same questions here, but suggestions like use new version of qiime or appropriate classifier dint work for me.
I used every silva classifier from here https://resources.qiime2.org/
Still no results

Is this have something to do with the scikit-learn installed (1.4.2) that is a piece of python?

I will appreciate any ideas
Thank you

Hi @ShevchenkoAlla,

The issue lies in the version of scikit-learn that's associated with the version of QIIME 2 that was used to generate the artifact you're using as input. Put a different way, the version of scikit-learn that's used to generate a particular artifact must match the current version of scikit-learn in the environment you're using. My guess is that the artifact you're using as input is from an older version of QIIME 2 that has a different version of scikit-learn (we only just upgraded to 1.4.2 in the 2024.5 release).

To resolve this issue, you'd need to either re-generate this artifact using the current version of QIIME 2, or run feature-classifier on an older version of QIIME 2 that is using scikit-learn 0.20.2 (which looks like would be a version from 2019, so it would be quite old).

Cheers :lizard:

2 Likes

Thank you so much!
I figured it out.

Best

1 Like