Does feature-classifier use the same kmer length for training and classification?

[DKioroglou (Dimitrios Kioroglou)]

I would like to ask the k-mer value qiime (version 2018.2) uses with classify-sklearn. For the training process the default value is 7. Is 7 also the value used during the classification process? Or is it 8 a mentioned?

Thank you in advance.

[Nicholas_Bokulich]

@DKioroglou

For the training process the default value is 7. Is 7 also the value used during the classification process?

The classifier always uses the same feature extraction pipeline that was used for training. So if you set it to 7 during training, it will use 7 for classification.