Processor.py

import cPickle as pickle
import metis_graph as mg
import matplotlib.pyplot as plt
import guide_hair as gh
import crash_on_ipy
import mcimport
from progressbar import *
import numpy as np
import weight_estimate as wet
import time

n_step = 100
n_group = 50
n_frame = 20
prefix = ''
file1 = "E:/cache/329.xml"
file2 = "../../maya cache/03074/hair_nRigidShape1.xml"
fileName = file2

np.set_printoptions(suppress=True)

print "training start at:", \
	time.strftime('%Y-%m-%d  %Hh%Mm%Ss',time.localtime(time.time()))
starttime = time.strftime('%Y-%m-%d  %Hh%Mm%Ss',time.localtime(time.time()))

particle_graph = pickle.load(file(prefix+'mgB.test'))
print "load mgB!"

import pymetis
_g = particle_graph
cut, vers = pymetis.part_graph(n_group, xadj=_g.xadj, adjncy=_g.adjncy, eweights=_g.eweights)

for i in range(n_group):
    print "group %d: %d strands"%(i, vers.count(i))

hairGroup = gh.GroupedGraph(particle_graph, vers)
hairGroup.solve()

frames = mcimport.importFile(fileName, n_frame)

count = 0
print "computing motion matrix of guides..."
pbar = ProgressBar().start()
for frame in frames:
    frame.loadCache(file(".dump/"+prefix+"frame"+str(count)+".dump", 'rb'))
    frame.calcSelectedParticleMotionMatrices(frames[0], hairGroup.guide)
    pbar.update((count/(len(frames)-1.0))*100)
    count += 1
pbar.finish()

model = wet.SkinModel(frames, hairGroup)
model.estimate()

t = str(time.time())
model.dump(file(".dump/"+prefix+t+"weights.dump", 'wb'))

print "training end at:", \
	time.strftime('%Y-%m-%d  %Hh%Mm%Ss',time.localtime(time.time()))

endtime = time.strftime('%Y-%m-%d  %Hh%Mm%Ss',time.localtime(time.time()))
from sendMail import *

content = 'graph generated!\nfrom '+starttime+' to '+endtime+'\n'
content += 'group %d, prefix %s\n' % (n_group, prefix)
content += 'guide sum %f, energy from %f t0 %f\n' % (hairGroup.energy, model.error0, model.error)

f = open("./Report/"+t+".txt", 'w');
f.write(content)
f.close()

send_mail(mailto_list, 'Report in 30/March '+prefix, content)