Mercurial > repos > recetox > aplcms_to_ramclustr_converter

annotate aplcms_to_ramclustr_converter.py @ 0:c6c0f6027e34 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/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
author recetox
date Tue, 22 Mar 2022 16:06:28 +0000
parents
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
1 #!/usr/bin/env python
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
2
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
3 import argparse
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
4 import sys
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
5
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
6 import pandas as pd
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
7
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
8
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
9 parser = argparse.ArgumentParser()
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
10 parser.add_argument("--dataframe", help="Parquet dataframe")
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
11 parser.add_argument('output')
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
12 args = parser.parse_args()
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
13
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
14
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
15 def main():
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
16 featureTable = pd.read_parquet(args.dataframe)
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
17
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
18 # Concatenate "mz" and "rt" columns; select relevant columns; pivot the table
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
19 featureTable["mz_rt"] = featureTable["mz"].astype(str) + "_" + featureTable["rt"].astype(str)
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
20 featureTable = featureTable[["sample", "mz_rt", "sample_intensity"]]
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
21 featureTable = pd.pivot_table(featureTable, columns="mz_rt", index="sample", values="sample_intensity")
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
22
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
23 try:
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
24 featureTable.to_csv(args.output, sep=',')
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
25 msg = f"Dataset of {len(featureTable)} samples is converted to a feature-by-sample table"
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
26 print(msg, file=sys.stdout)
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
27 return 0
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
28 except Exception:
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
29 print("Could not write the data", file=sys.stdout)
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
30 return 1
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
31
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
32
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
33 if __name__ == "__main__":
c6c0f6027e34 "planemo upload for repository https://github.com/RECETOX/galaxytools/tools/aplcms_to_ramclustr_converter/ commit 4d2ac914c951166e386a94d8ebb8cb1becfac122"
recetox
parents:
diff changeset
34 main()