Mercurial > repos > bgruening > sklearn_model_validation
diff model_prediction.py @ 24:a5aed87b2cc0 draft
"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/sklearn commit 5b2ac730ec6d3b762faa9034eddd19ad1b347476"
author | bgruening |
---|---|
date | Mon, 16 Dec 2019 05:28:32 -0500 |
parents | 5895fe0b8bde |
children | 9b017b0da56e |
line wrap: on
line diff
--- a/model_prediction.py Thu Nov 07 05:35:17 2019 -0500 +++ b/model_prediction.py Mon Dec 16 05:28:32 2019 -0500 @@ -2,13 +2,11 @@ import json import numpy as np import pandas as pd -import tabix import warnings from scipy.io import mmread from sklearn.pipeline import Pipeline -from galaxy_ml.externals.selene_sdk.sequences import Genome from galaxy_ml.utils import (load_model, read_columns, get_module, try_get_attr) @@ -138,45 +136,10 @@ pred_data_generator = klass( ref_genome_path=ref_seq, vcf_path=vcf_path, **options) - pred_data_generator.fit() + pred_data_generator.set_processing_attrs() variants = pred_data_generator.variants - # TODO : remove the following block after galaxy-ml v0.7.13 - blacklist_tabix = getattr(pred_data_generator.reference_genome_, - '_blacklist_tabix', None) - clean_variants = [] - if blacklist_tabix: - start_radius = pred_data_generator.start_radius_ - end_radius = pred_data_generator.end_radius_ - for chrom, pos, name, ref, alt, strand in variants: - center = pos + len(ref) // 2 - start = center - start_radius - end = center + end_radius - - if isinstance(pred_data_generator.reference_genome_, Genome): - if "chr" not in chrom: - chrom = "chr" + chrom - if "MT" in chrom: - chrom = chrom[:-1] - try: - rows = blacklist_tabix.query(chrom, start, end) - found = 0 - for row in rows: - found = 1 - break - if found: - continue - except tabix.TabixError: - pass - - clean_variants.append((chrom, pos, name, ref, alt, strand)) - else: - clean_variants = variants - - setattr(pred_data_generator, 'variants', clean_variants) - - variants = np.array(clean_variants) # predict 1600 sample at once then write to file gen_flow = pred_data_generator.flow(batch_size=1600)