DAMASK_EICMD/processing/post/permuteData.py

114 lines
4.1 KiB
Python
Raw Normal View History

#!/usr/bin/env python3
import os
import sys
from optparse import OptionParser
import numpy as np
import damask
scriptName = os.path.splitext(os.path.basename(__file__))[0]
scriptID = ' '.join([scriptName,damask.version])
# --------------------------------------------------------------------
# MAIN
# --------------------------------------------------------------------
2019-02-16 22:55:41 +05:30
parser = OptionParser(option_class=damask.extendableOption, usage='%prog options [ASCIItable(s)]', description = """
Permute all values in given column(s).
""", version = scriptID)
parser.add_option('-l','--label',
dest = 'label',
action = 'extend', metavar = '<string LIST>',
help ='column(s) to permute')
parser.add_option('-u', '--unique',
dest = 'unique',
action = 'store_true',
help = 'shuffle unique values as group')
parser.add_option('-r', '--rnd',
dest = 'randomSeed',
type = 'int', metavar = 'int',
help = 'seed of random number generator [%default]')
parser.set_defaults(label = [],
2020-01-26 19:44:16 +05:30
unique = False,
randomSeed = None,
)
(options,filenames) = parser.parse_args()
if len(options.label) == 0:
parser.error('no labels specified.')
# --- loop over input files -------------------------------------------------------------------------
2015-08-18 20:07:32 +05:30
if filenames == []: filenames = [None]
for name in filenames:
2015-08-18 20:07:32 +05:30
try:
2020-02-20 19:35:38 +05:30
table = damask.ASCIItable(name = name)
2020-01-29 04:09:46 +05:30
except IOError:
continue
damask.util.report(scriptName,name)
# ------------------------------------------ read header ------------------------------------------
table.head_read()
# ------------------------------------------ process labels ---------------------------------------
errors = []
remarks = []
columns = []
dims = []
indices = table.label_index (options.label)
dimensions = table.label_dimension(options.label)
for i,index in enumerate(indices):
if index == -1: remarks.append('label "{}" not present...'.format(options.label[i]))
else:
columns.append(index)
dims.append(dimensions[i])
if remarks != []: damask.util.croak(remarks)
if errors != []:
damask.util.croak(errors)
table.close(dismiss = True)
continue
# ------------------------------------------ assemble header ---------------------------------------
2020-02-21 15:30:53 +05:30
randomSeed = int(os.urandom(4).hex(), 16) if options.randomSeed is None else options.randomSeed # random seed per file
np.random.seed(randomSeed)
table.info_append([scriptID + '\t' + ' '.join(sys.argv[1:]),
'random seed {}'.format(randomSeed),
])
table.head_write()
# ------------------------------------------ process data ------------------------------------------
table.data_readArray() # read all data at once
for col,dim in zip(columns,dims):
if options.unique:
s = set(map(tuple,table.data[:,col:col+dim])) # generate set of (unique) values
uniques = np.array(map(np.array,s)) # translate set to np.array
shuffler = dict(zip(s,np.random.permutation(len(s)))) # random permutation
table.data[:,col:col+dim] = uniques[np.array(map(lambda x: shuffler[tuple(x)],
table.data[:,col:col+dim]))] # fill table with mapped uniques
else:
np.random.shuffle(table.data[:,col:col+dim]) # independently shuffle every row
# ------------------------------------------ output result -----------------------------------------
table.data_writeArray()
# ------------------------------------------ output finalization -----------------------------------
table.close() # close ASCII tables