def calculateOneStation(id): item0 = list[id] station_0 = list_station_all[id] for x in xrange(1,size): print x item = list_station_all[x] leng = len(item) #print leng #cov =np.cov(station_0,item)[0][1] #correlate = np.corrcoef(station_0, item)[0,1] semi = semivariance(station_0, item) #list_corr_tuple.append((id, x, correlate)) list_corr_tuple.append((id, x, semi)) #print correlate #correlate = corr(station_0, item) #value = 0 #for y in xrange(0, leng): # value = value + abs(station_0[y] - item[y]) #value = value/leng #print value list_index.append(x) list_value.append(semi) list_h.append(haversine(item0[2], item0[1], list[x][2], list[x][1]))
def generateGlobalMatrix( date_from, date_to, list_all_station):
global trend_curve
list_station = list_all_station
#print list_station
meteo_size = len(list_station)
list_mean = []
for i in xrange(0,meteo_size):
#print list_station[i][0]
item = getTempOfAAtationFromTo( list_station[i][3],date_from, date_to)
list_all = zip(*item)
#3 for temp, 4 for humid, 5 for rain
list_temp = list_all[3]
#print list_temp
list_data.append(list_temp)
mean_item = (list_all[1][0],list_all[2][0], sum(list_all[3])/len(list_all[3]))
print mean_item
list_mean.append(mean_item)
f.write('MEAN: ' + str(list_mean) + '\n')
trend_curve = curveFitting(list_mean)
print trend_curve
#plotSurface(trend_curve, list_mean)
#print len(list_mean)
#print list_data[0]
#G = np.zeros([meteo_size, meteo_size])
for x in xrange(0,meteo_size):#x 0->97 is Meteo_id 1->98
for y in xrange(0,meteo_size):#0->97 is Meteo_id 1->98
#gen covariance for [x][y]
a = list_data[x]
b = list_data[y]
#cov =np.cov(a,b)[0][1]
semi = semivariance( a, b)
cov = covariance(a,b)
#print 'a: ', a
#print 'b: ',b
#print 'cov a b = ', str(cov)
#print covariance(a,b)
G[x][y]= semi
G_COV[x][y] = cov
'''
f.write('G_SEMI ----------------------------------------- \n')
for item in G:
f.write(str(item) + '\n')
f.write('G_COV ----------------------------------------- \n')
for item in G_COV:
f.write(str(item) + '\n')
'''
#print G[0:5,0:5]
return trend_curve
