gtnet.filter module

gtnet.filter.get_cutoffs(rocs, fpr)

Get score cutoffs to achieve desired false-positive rate

Parameters:

  • rocs (dict) – The ROC curves for each taxonomic level

  • fpr (float) – The false-positive rate to get the score for

gtnet.filter.filter(argv=None)

Filter raw taxonomic classifications

gtnet.filter.filter_predictions(pred_df, cutoffs)

Filter taxonomic classification predictions

Parameters:

  • pred_df (DataFrame) – The DataFrame containing predictions and confidence scores for each taxonomic level

  • cutoffs (dict) – A dictionary containing the confidence score cutoff for each taxonomic level