Mercurial > repos > immport-devteam > flowclr_summary

annotate flowclrstats.py @ 1:7a889f2f2e15 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/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
author azomics
date Mon, 22 Jun 2020 19:55:57 -0400
parents
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
1
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
1 #!/usr/bin/env python
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
2
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
3 ######################################################################
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
4 # Copyright (c) 2016 Northrop Grumman.
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
5 # All rights reserved.
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
6 ######################################################################
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
7
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
8 from __future__ import print_function
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
9 from argparse import ArgumentParser
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
10 import pandas as pd
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
11
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
12
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
13 def get_FLOCK_stats(input_file, output_file, out_file2):
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
14 df = pd.read_table(input_file)
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
15 summary = df.groupby('Population').describe().round(1)
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
16 counts = df['Population'].value_counts()
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
17 percent = (df['Population'].value_counts(normalize=True) * 100).round(decimals=2)
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
18 tot_count = len(df['Population'])
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
19
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
20 to_rm = summary.loc(axis=0)[:, ['count']].index.tolist()
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
21 df1 = summary[~summary.index.isin(to_rm)]
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
22 df1.to_csv(out_file2, sep="\t")
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
23
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
24 with open(output_file, "w") as outf:
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
25 outf.write("Population\tCount\tPercentage\n")
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
26 for pops in set(df.Population):
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
27 outf.write("\t".join([str(pops), str(counts.loc[pops]), str(percent.loc[pops])]) + "\n")
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
28 outf.write("Total\t" + str(tot_count) + "\t \n")
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
29 return
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
30
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
31
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
32 if __name__ == '__main__':
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
33 parser = ArgumentParser(
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
34 prog="flowstats",
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
35 description="Gets statistics on FLOCK run")
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
36
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
37 parser.add_argument(
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
38 '-i',
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
39 dest="input_file",
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
40 required=True,
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
41 help="File locations for flow clr file.")
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
42
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
43 parser.add_argument(
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
44 '-o',
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
45 dest="out_file",
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
46 required=True,
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
47 help="Path to the directory for the output file.")
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
48
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
49 parser.add_argument(
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
50 '-p',
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
51 dest="out_file2",
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
52 required=True,
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
53 help="Path to the directory for the output file.")
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
54 args = parser.parse_args()
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
55
7a889f2f2e15 "planemo upload for repository https://github.com/ImmPortDB/immport-galaxy-tools/tree/master/flowtools/flowclr_summary commit ce895377ed593ace77016bd019a7998e13e470cc"
azomics
parents:
diff changeset
56 get_FLOCK_stats(args.input_file, args.out_file, args.out_file2)