223031012
/
DAMASK_EICMD
mirror of https://github.com/achalhp/DAMASK_EICMD.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

added group shuffling (i.e. exchanging unique values)

This commit is contained in:
Philip Eisenlohr 2015-12-04 01:17:45 +00:00
parent 08014a7030
commit 937634d703
1 changed files with 14 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
processing/post/permuteData.py
View File

@ -22,12 +22,17 @@ 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 = [],
unnique = False,
randomSeed = None,
)
@ -61,7 +66,7 @@ for name in filenames:
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]))
if index == -1: remarks.append('label "{}" not present...'.format(options.label[i]))
else:
columns.append(index)
dims.append(dimensions[i])
@ -86,7 +91,14 @@ for name in filenames:
table.data_readArray() # read all data at once
for col,dim in zip(columns,dims):
table.data[:,col:col+dim] = np.random.permutation(table.data[:,col:col+dim])
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 -----------------------------------------