Mercurial > repos > jay > pdaug_word_vector_model

annotate PDAUG_AddClassLabel/PDAUG_AddClassLabel.py @ 10:bc1744905ef4 draft default tip

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 d396d7ff89705cc0dd626ed32c45a9f4029b1b05"
author jay
date Wed, 12 Jan 2022 20:17:55 +0000
parents 3ce435b8d648
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
1 import pandas as pd
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
2 import argparse
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
3
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
4 parser = argparse.ArgumentParser()
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
5
3ce435b8d648 "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")
3ce435b8d648 "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")
3ce435b8d648 "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)")
3ce435b8d648 "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")
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
10
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
11 args = parser.parse_args()
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
12
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
13 df1 = pd.read_csv(args.InFile, sep="\t")
3ce435b8d648 "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])
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
15
3ce435b8d648 "planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff changeset
16 df = pd.concat([df1, df2], axis=1)
3ce435b8d648 "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)