def makePatchStatsDict(patchDict, minpatchDataDict):
zoneDict = minpatchDataDict["zoneDictionary"]
patchStatsDictionary = {}
allAreaList, validAreaList = makePatchAreaLists(patchDict, zoneDict)
if len(allAreaList) > 0:
medianAllPatch = stats.lmedianscore(allAreaList)
else:
medianAllPatch = 0
if len(validAreaList) > 0:
medianValidPatch = stats.lmedianscore(validAreaList)
else:
medianValidPatch = 0
patchStatsDictionary["AllPatchCount"] = len(allAreaList)
patchStatsDictionary["AllPatchArea"] = sum(allAreaList)
patchStatsDictionary["medianAllPatch"] = medianAllPatch
patchStatsDictionary["ValidPatchCount"] = len(validAreaList)
patchStatsDictionary["ValidPatchArea"] = sum(validAreaList)
patchStatsDictionary["medianValidPatch"] = medianValidPatch
return patchStatsDictionary
def makePatchStatsDict(patchDict, minpatchDataDict):
zoneDict = minpatchDataDict["zoneDictionary"]
patchStatsDictionary = {}
allAreaList, validAreaList = makePatchAreaLists(patchDict, zoneDict)
if len(allAreaList) > 0:
medianAllPatch = stats.lmedianscore(allAreaList)
else:
medianAllPatch = 0
if len(validAreaList) > 0:
medianValidPatch = stats.lmedianscore(validAreaList)
else:
medianValidPatch = 0
patchStatsDictionary["AllPatchCount"] = len(allAreaList)
patchStatsDictionary["AllPatchArea"] = sum(allAreaList)
patchStatsDictionary["medianAllPatch"] = medianAllPatch
patchStatsDictionary["ValidPatchCount"] = len(validAreaList)
patchStatsDictionary["ValidPatchArea"] = sum(validAreaList)
patchStatsDictionary["medianValidPatch"] = medianValidPatch
return patchStatsDictionary
def estimate(self,l):
def sumvar(l,m):
return stats.sum(map(lambda x:(x-m)**2,l))
med=stats.lmedianscore(l)
oldmed=med+1
while med!=oldmed:
l0=filter(lambda x: x<med,l)
l1=filter(lambda x: x>med,l)
lm=filter(lambda x: x==med,l)
mu0=stats.mean(l0*2+lm)
mu1=stats.mean(l1*2+lm)
(oldmed,med)=(med,(mu0+mu1)/2)
#print mu0,mu1, oldmed,med#,l0,l1
self.sigma=max(0.1**10, math.sqrt((sumvar(l0+lm,mu0)+sumvar(l1,mu1)+(mu1-med)**2)/(len(l))))
self.p0=(len(l0)+len(lm)*0.5)/len(l)
self.p1=(len(l1)+len(lm)*0.5)/len(l)
self.mu0=mu0
self.mu1=mu1
return self.mu0, self.mu1, self.sigma, self.p0, self.p1
def estimate(self, l):
def sumvar(l, m):
return stats.sum(map(lambda x: (x - m)**2, l))
med = stats.lmedianscore(l)
oldmed = med + 1
while med != oldmed:
l0 = filter(lambda x: x < med, l)
l1 = filter(lambda x: x > med, l)
lm = filter(lambda x: x == med, l)
mu0 = stats.mean(l0 * 2 + lm)
mu1 = stats.mean(l1 * 2 + lm)
(oldmed, med) = (med, (mu0 + mu1) / 2)
#print mu0,mu1, oldmed,med#,l0,l1
self.sigma = max(
0.1**10,
math.sqrt((sumvar(l0 + lm, mu0) + sumvar(l1, mu1) +
(mu1 - med)**2) / (len(l))))
self.p0 = (len(l0) + len(lm) * 0.5) / len(l)
self.p1 = (len(l1) + len(lm) * 0.5) / len(l)
self.mu0 = mu0
self.mu1 = mu1
return self.mu0, self.mu1, self.sigma, self.p0, self.p1
