Mercurial > repos > jay > pdaug_peptide_ngrams

annotate PDAUG_AddClassLabel/PDAUG_AddClassLabel.py @ 2:bc6aef6ec2d4 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit 8b18552f6d2b2261efebe1075ff4c18a295b94dd"
author jay
date Tue, 29 Dec 2020 18:38:48 +0000
parents 7557b48b2872
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
1 import pandas as pd
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
2 import argparse
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
3
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
4 parser = argparse.ArgumentParser()
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
5
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
6 parser.add_argument("-I", "--InFile", required=True, default=None, help="Input data file")
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
7 parser.add_argument("-C", "--ClassLabel", required=False, default=0, help="Class Label 0 or 1")
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
8 parser.add_argument("-t", "--ClassLabelTitle", required=False, default='Class_label', help="Title to use for class label column (Class_label)")
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
9 parser.add_argument("-O", "--OutFile", required=False, default='OutFile.tsv', help="Output file name")
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
10
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
11 args = parser.parse_args()
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
12
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
13 df1 = pd.read_csv(args.InFile, sep="\t")
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
14 df2 = pd.DataFrame([args.ClassLabel]*df1.shape[0], columns=[args.ClassLabelTitle])
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
15
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
16 df = pd.concat([df1, df2], axis=1)
7557b48b2872 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
17 df.to_csv(args.OutFile, sep="\t", index=False)