1#!/usr/bin/env python3 

2 

3import os 

4from argparse import ArgumentParser, RawDescriptionHelpFormatter 

5from glob import glob 

6from llama.dev.background.pvalue import __doc__ 

7 

8OUTFILE = 'populations.hdf5' 

9 

10 

11def get_parser(): 

12 """Get CLI parser.""" 

13 parser = ArgumentParser(formatter_class=RawDescriptionHelpFormatter, 

14 description=__doc__.replace(".. code:\n\n", "")) 

15 parser.add_argument("-c", "--clobber", action="store_true", help=f""" 

16 Overwrite "./{OUTFILE}" if it already exists. Will not overwrite by 

17 default.""") 

18 return parser 

19 

20 

21def main(): 

22 """Run CLI.""" 

23 import h5py 

24 import pandas 

25 import numpy as np 

26 parser = get_parser() 

27 args = parser.parse_args() 

28 tables = glob('*.txt') 

29 if os.path.isfile(OUTFILE) and not args.clobber: 

30 print(f"{OUTFILE} already exists, not overwriting; QUITTING.") 

31 exit(1) 

32 with h5py.File(OUTFILE, 'w') as outfile: 

33 for table in tables: 

34 path = table.split('.')[0].replace('-', '/') 

35 dfrm = pandas.read_csv(table, skiprows=(1,), delim_whitespace=True) 

36 # not worth using compression here; savings minimal. method here: 

37 # http://docs.h5py.org/en/stable/high/dataset.html#filter-pipeline 

38 outfile[path] = np.sort(dfrm['EVENT-ODDS-RATIO'], kind='heapsort') 

39 

40 

41if __name__ == "__main__": 

42 main()