def pearsonfit(data):
data=np.array(data)
nozero=len(data.nonzero()[0])
pze=1-float(nozero)/len(data)
para=lm.pelpe3(lm.samlmu(data[data!=0],3))
p3= np.array([lm.cdfpe3(i,para) for i in data])
p3=stats.norm.ppf(p3)
return p3
def pearsonfit(data):
data = np.array(data)
nozero = len(data.nonzero()[0])
pze = 1 - float(nozero) / len(data)
#print pze
#data2=ma.masked_values(data.values, 0)
para = lm.pelpe3(lm.samlmu(data[data != 0], 3))
p3 = np.array([lm.cdfpe3(i, para) for i in data])
p3 = stats.norm.ppf(p3)
#p3=stats.norm.ppf(pze+(1-pze)*lm.cdfnor(p3,[0,1]))
return p3
comparefunc(norlmom,correctnorlmom,"LMOMNOR",6)
#RANDNOR
try:
norrand = lmoments.randnor(10,correctnorfit)
if len(norrand) != 10:
print("RANDNOR FAILED")
except:
print("RANDNOR FAILED")
print("#######################################")
#######################################
##PE3
#######################################
##PELPE3
pe3fit = lmoments.pelpe3(LMU)
correctpe3fit = [3.236364, 2.199489, 1.646184]
comparefunc(pe3fit,correctpe3fit,"PELPE3",6)
##QUAPE3
pe3qua = [lmoments.quape3(0.2,correctpe3fit),lmoments.quape3(0.5,correctpe3fit),lmoments.quape3(0.8,correctpe3fit)]
pe3qua2 = lmoments.quape3([0.2,0.5,0.8],correctpe3fit)
correctpe3qua = [1.447672, 2.663015, 4.705896]
comparefunc(pe3qua,correctpe3qua,"QUAPE3",6)
comparefunc(pe3qua2,correctpe3qua,"QUAPE3 group",6)
##LMRPE3
pe3lmr = lmoments.lmrpe3(correctpe3fit,4)
correctpe3lmr = [3.2363636, 1.1418182, 0.2738845, 0.1498865]
comparefunc(pe3lmr,correctpe3lmr,"LMRPE3",6)
# 3 PE3:
# para1 = \para1
# para2 = \para2 [para2=xmom[0]/alpha]
# para3 = \gamma
if FUNC == 'PE3':
#for i in progressbar(range(ysize), "Computing: ", 40):
for i in range(ysize):
if i % 10 == 0: print(i, 'out of ', ysize)
for j in range(xsize):
if np.nanmean(datm[:, i, j]) > -9990.:
# There are many grids with constant values or
# there is only one large value but others are constant.
# We cannot calculate the parameters with this time series.
if np.std(datm[:10, i, j]) > 1e-5:
lmoms = lmom.samlmu(datm[:, i, j], 4)
params = lmom.pelpe3(lmoms)
try:
para1[i, j] = params[0]
para2[i, j] = params[1]
para3[i, j] = params[2]
p_AIC[i, j] = lmom.AIC(datm[:, i, j], FUNC)
y = lmom.quape3(p, params)
c_AIC[i, j] = calc_aic(datm[:, i, j], y)
py_AIC[i, j] = aic.aic(datm[:, i, j], y, len(params))
except:
print("error in calculating the params", i, j, params)
print("datm:", datm[:, i, j])
print(
"daty",
lmom.quape3(
p, [para1[i, j], para2[i, j], para3[i, j]]))
def _pe3(self, dat, l):
''' Fit a Pearson type 3 distribution'''
para = lm.pelpe3(lm.samlmu(dat, l))
_cdf = partial(lm.cdfpe3, para=para)
_pdf = partial(lm.pdfpe3, para=para)
return para, _cdf, _pdf