DAMASK_EICMD/processing/legacy/addEuclideanDistance.py

188 lines
8.4 KiB
Python
Raw Normal View History

#!/usr/bin/env python3
import os
import sys
2019-12-21 23:34:29 +05:30
from io import StringIO
from optparse import OptionParser
import itertools
import numpy as np
from scipy import ndimage
import damask
scriptName = os.path.splitext(os.path.basename(__file__))[0]
scriptID = ' '.join([scriptName,damask.version])
def periodic_3Dpad(array, rimdim=(1,1,1)):
rimdim = np.array(rimdim,'i')
size = np.array(array.shape,'i')
padded = np.empty(size+2*rimdim,array.dtype)
padded[rimdim[0]:rimdim[0]+size[0],
rimdim[1]:rimdim[1]+size[1],
rimdim[2]:rimdim[2]+size[2]] = array
p = np.zeros(3,'i')
2016-10-25 00:46:29 +05:30
for side in range(3):
for p[(side+2)%3] in range(padded.shape[(side+2)%3]):
for p[(side+1)%3] in range(padded.shape[(side+1)%3]):
for p[side%3] in range(rimdim[side%3]):
spot = (p-rimdim)%size
padded[p[0],p[1],p[2]] = array[spot[0],spot[1],spot[2]]
2016-10-25 00:46:29 +05:30
for p[side%3] in range(rimdim[side%3]+size[side%3],size[side%3]+2*rimdim[side%3]):
spot = (p-rimdim)%size
padded[p[0],p[1],p[2]] = array[spot[0],spot[1],spot[2]]
return padded
# --------------------------------------------------------------------
# MAIN
# --------------------------------------------------------------------
features = [
{'aliens': 1, 'names': ['boundary','biplane'],},
{'aliens': 2, 'names': ['tripleline',],},
{'aliens': 3, 'names': ['quadruplepoint',],}
]
neighborhoods = {
'neumann':np.array([
[-1, 0, 0],
[ 1, 0, 0],
[ 0,-1, 0],
[ 0, 1, 0],
[ 0, 0,-1],
[ 0, 0, 1],
]),
'moore':np.array([
[-1,-1,-1],
[ 0,-1,-1],
[ 1,-1,-1],
[-1, 0,-1],
[ 0, 0,-1],
[ 1, 0,-1],
[-1, 1,-1],
[ 0, 1,-1],
[ 1, 1,-1],
#
[-1,-1, 0],
[ 0,-1, 0],
[ 1,-1, 0],
[-1, 0, 0],
#
[ 1, 0, 0],
[-1, 1, 0],
[ 0, 1, 0],
[ 1, 1, 0],
#
[-1,-1, 1],
[ 0,-1, 1],
[ 1,-1, 1],
[-1, 0, 1],
[ 0, 0, 1],
[ 1, 0, 1],
[-1, 1, 1],
[ 0, 1, 1],
[ 1, 1, 1],
])
}
2019-02-16 22:11:56 +05:30
parser = OptionParser(option_class=damask.extendableOption, usage='%prog options [ASCIItable(s)]', description = """
Add column(s) containing Euclidean distance to grain structural features: boundaries, triple lines, and quadruple points.
""", version = scriptID)
parser.add_option('-p',
'--pos', '--position',
dest = 'pos', metavar = 'string',
help = 'label of coordinates [%default]')
parser.add_option('-i',
'--id', '--identifier',
dest = 'id', metavar = 'string',
help='label of grain identifier [%default]')
parser.add_option('-t',
'--type',
dest = 'type', action = 'extend', metavar = '<string LIST>',
help = 'feature type {{{}}} '.format(', '.join(map(lambda x:'/'.join(x['names']),features))) )
parser.add_option('-n',
'--neighborhood',
dest = 'neighborhood', choices = list(neighborhoods.keys()), metavar = 'string',
help = 'neighborhood type [neumann] {{{}}}'.format(', '.join(neighborhoods.keys())))
parser.add_option('-s',
'--scale',
dest = 'scale', type = 'float', metavar = 'float',
help = 'voxel size [%default]')
parser.set_defaults(pos = 'pos',
id = 'texture',
neighborhood = 'neumann',
scale = 1.0,
)
(options,filenames) = parser.parse_args()
if filenames == []: filenames = [None]
if options.type is None:
parser.error('no feature type selected.')
if not set(options.type).issubset(set(list(itertools.chain(*map(lambda x: x['names'],features))))):
parser.error('type must be chosen from (%s).'%(', '.join(map(lambda x:'|'.join(x['names']),features))) )
if 'biplane' in options.type and 'boundary' in options.type:
parser.error('only one from aliases "biplane" and "boundary" possible.')
feature_list = []
for i,feature in enumerate(features):
for name in feature['names']:
for myType in options.type:
if name.startswith(myType):
feature_list.append(i) # remember valid features
break
for name in filenames:
damask.util.report(scriptName,name)
table = damask.Table.load(StringIO(''.join(sys.stdin.read())) if name is None else name)
2020-12-08 05:06:41 +05:30
grid,size,origin = damask.grid_filters.cellsSizeOrigin_coordinates0_point(table.get(options.pos))
neighborhood = neighborhoods[options.neighborhood]
diffToNeighbor = np.empty(list(grid+2)+[len(neighborhood)],'i')
microstructure = periodic_3Dpad(table.get(options.id).astype('i').reshape(grid,order='F'))
for i,p in enumerate(neighborhood):
stencil = np.zeros((3,3,3),'i')
stencil[1,1,1] = -1
stencil[p[0]+1,
p[1]+1,
p[2]+1] = 1
diffToNeighbor[:,:,:,i] = ndimage.convolve(microstructure,stencil) # compare ID at each point...
# ...to every one in the specified neighborhood
# for same IDs at both locations ==> 0
diffToNeighbor = np.sort(diffToNeighbor) # sort diff such that number of changes in diff (steps)...
# ...reflects number of unique neighbors
uniques = np.where(diffToNeighbor[1:-1,1:-1,1:-1,0] != 0, 1,0) # initialize unique value counter (exclude myself [= 0])
for i in range(1,len(neighborhood)): # check remaining points in neighborhood
uniques += np.where(np.logical_and(
diffToNeighbor[1:-1,1:-1,1:-1,i] != 0, # not myself?
diffToNeighbor[1:-1,1:-1,1:-1,i] != diffToNeighbor[1:-1,1:-1,1:-1,i-1],
), # flip of ID difference detected?
1,0) # count that flip
distance = np.ones((len(feature_list),grid[0],grid[1],grid[2]),'d')
for i,feature_id in enumerate(feature_list):
distance[i,:,:,:] = np.where(uniques >= features[feature_id]['aliens'],0.0,1.0) # seed with 0.0 when enough unique neighbor IDs are present
distance[i,:,:,:] = ndimage.morphology.distance_transform_edt(distance[i,:,:,:])*[options.scale]*3
distance = distance.reshape([len(feature_list),grid.prod(),1],order='F')
for i,feature in enumerate(feature_list):
2020-09-14 10:34:01 +05:30
table = table.add('ED_{}({})'.format(features[feature]['names'][0],options.id),
distance[i,:],
scriptID+' '+' '.join(sys.argv[1:]))
2020-11-24 01:24:07 +05:30
table.save((sys.stdout if name is None else name))