# -*- coding: utf-8 -*-

import time,re

import h5py,re,dask,os,sys

import numpy as np

import dask.array as da

import dask.dataframe as dd

import time

import progressbar

glist = [eval(i.strip()) for i in tuple(open('centrality/newsubgroups.txt'))]

# dtol = 90

#glist = filter(lambda x:int(x[1])>tol,glist)

inorganics = ''.join(open('src/background/inorganic_mcm.kpp','r').readlines())

inorganics = re.findall(r'\b([\w\d]+)\s*=\s*IGNORE',inorganics)

inorganics.append('CO')

flist = [ filter(lambda x: x not in inorganics, i) for i in glist]

flist = filter(lambda x: len(x)>1,flist)

print len(flist), 'clusters'

#######

fl = 'lhs.h5'

maxtsps = int(144/2.)*3

desired = 24

tslst = range(1,(3*144)+1,6)#maxtsps/desired)

setsgroup= 4

run = True

data = [[]]*len(flist)

bar = progressbar.ProgressBar()

with h5py.File(fl,'r') as hf:

groups = list(filter(lambda x: type(x[1])==h5py._hl.group.Group, hf.items()))

print 'every '+str(setsgroup)+' groups'

for gr in bar(groups[::setsgroup]):

g=gr[1]

if run:

shead = g.attrs['spechead'].split(',')

fhead = g.attrs['fluxhead'].split(',')

spec = dd.from_array(g.get('spec')[1:,:],chunksize=50000, columns = shead)

flux = dd.from_array(g.get('flux')[1:,:],chunksize=50000,columns = fhead)

if run:

#spec = spec.set_index('TIME', sorted=True)

M = spec.M.mean()

spec = spec/M

#flux = flux.set_index('TIME', sorted=True)

shead = g.attrs['spechead'].split(',')

fcol = g.attrs['fluxhead']

products = [i.split('+') for i in re.findall(r'-->([A-z0-9+]*)',fcol)]

prodloss = {k: {'loss':[],'prod':[]} for k in shead}

### reaction prodloss arrays

for idx in xrange(len(products)):

for i in products[idx]:

try:prodloss[i]['prod'].append(idx)

except:None

run = False

print 'group ',g

for gn,g in enumerate(flist):

sm = (spec.loc[tslst,g]).sum(axis=1).compute()*M

fx = [flux.loc[:,flux.columns[prodloss[s]['prod']]].sum(axis=1) for s in g]

sm = sum(fx)

for i,s in enumerate(g):

frac = fx[i]/sm

frac = frac.loc[tslst].compute()

std = frac.std()

mean = frac.mean()

data[gn].append([frac,std,mean])

#print g,gn

np.save('lhsgroupfract.npy',data)

fdsafdsaf=k

groups = eval(tuple(open('centrality/lhscollection.txt'))[2])

print groups[1]

a = new('cri22.h5')

maxtsps = len(ts)-144

desired = 24

tslst = range(144,len(ts),maxtsps/desired)

newts = ts[tslst]

percentstd = .05

print 'Percentage Tolerance %d%%'%(percentstd*100)

html = '''

'''

lumpfull =[]

grouplist=[]

groupcoeff=[]

M= M.compute()

for gn,g in enumerate(groups):

print g

#sm = np.log10(spec.loc[newts,g]).sum(axis=1)

sm = (spec.loc[newts,g]).sum(axis=1).compute()*M

html+='<h4> lmp %02d </h4>\n'%(gn+1)

dummygroup=[]

dummy = []

fx = [flux.loc[:,prod(s)].sum(axis=1) for s in g]

sm = sum(fx)

for i,s in enumerate(g):

#frac = np.log10(spec.loc[newts,s])/sm

frac = fx[i]/sm

frac = frac.loc[newts].compute()

std = frac.std()

mean = frac.mean()

inlim = ''

if std>percentstd:

print std,s,gn

inlim = 'style="color:powderblue;"'

line = list(' |'.join([str(int(i*4)) for i in list(frac)]))

dummy.append([s,float(mean),float(std)])

dummygroup.append(s)

groupcoeff.append([s,'%.08f'%(mean)])

count = 4

for c in (s+' %d%%'%(100*mean)).lower():

line[count] = c

count+=3

html += '<!--%s-->\n<p %s>{std:%02d percent} §%s {%02.1f °[-] }{%02.1f °[+]}</p>\n'%(s,inlim,100.*std,''.join(line),100.*frac.min(),100.*frac.max())

if inlim == '':

grouplist.append(dummygroup)

lumpfull.append(dummy)

html+='\n'

html+='</body>'

#print html

with open('test.html','w') as f:

f.write(html)

with open('lump.mech','w') as f:

f.write('lumplist = '+ str(grouplist).replace("'",'"'))

f.write('\nlumpcoeff = '+ str(groupcoeff).replace("'",'"'))