def oldstyle():
outf=open("report_2.txt", "w")
with open(os.environ["HOMEPATH"]+"\\Documents\\abrupt\\c_test\\x.x_08Jun",'r') as f:
lines=f.readlines()
for line in lines:
#print line
pos1=line.find('"')
if pos1>=0:
pos2=line.find('"',pos1+1)
#print pos1,pos2
fn=line[:pos1-1]
#print fn+":"
yset=eval(line[pos1:pos2+1])
#print "YSET", yset
_,yr,pre,post=line[pos2+1:].split()[:4]
# print yr,pre,post
dfn=fn[:-len('final.csv')-1]
tfn=os.environ["HOMEPATH"]+"\\Documents\\abrupt\\c_test\\"+dfn
#print fn, tfn
if dfn[:len("Global")]=="Global" or dfn[0] == 'r':
print "do",tfn
try:
direc=glob.glob(os.environ["HOMEPATH"]+"\\Documents\\abrupt\\c_test\\"+fn[:-len('final.csv')-1]+'/*.trace')[0]
bp=statbreaks.brkrpt(direc)
#print bp.years()
yrix=int(list(bp.years()).index(float(yr)))
preix=int(list(bp.years()).index(float(pre)))
postix=int(list(bp.years()).index(float(post)))
#print bp.years()[preix:postix+1]
ancova=doancova.ANCOVA(bp.years()[preix:postix+1], bp.ys()[preix:postix+1], bp.years()[yrix+1],Rpath='"C:\\Program Files\\R\\R-3.0.2\\bin\\Rscript.exe"')
ancova.parseResult(ancova.results())
print >>outf, "slope",line[:-1],":",ancova._ANCOVA__breakYear,ancova.firstSlope(),formatprobs(ancova.firstSlopeProb()), ancova.firstRsq(), ancova.secondSlope(),formatprobs(ancova.secondSlopeProb()),ancova.secondRsq(),ancova.shift(),formatprobs(ancova._ANCOVA__prDiffShift),":",
if dfn[:len("Global")]=="Global":
print >>outf, cmip3_info.matchfn(tfn),
elif dfn[0] == 'r':
print >>outf, cmip5_info.matchfn(tfn),
print >>outf, formatprobs(ancova.changeOfTrendProb(),ancova._ANCOVA__prDiffShift),formatprobs(ancova.changeOfTrendProb()*ancova._ANCOVA__prDiffShift)
#if
except Exception as e:
fl=os.environ["HOMEPATH"]+"\\Documents\\abrupt\\c_test\\"+fn[:-len('final.csv')-1]+'/*.trace'
print >>outf,"\n",line,"Error getting files",fl
print "using ", fl
#print glob.glob(fl)
#raise Exception(str(e))
pass
#print direc,
outf.close()
print "done"
def getannual(finalfn,datapath="C:\\Users\\s4493222\\Documents\\abrupt\\c_test\\"):
#assume *.final.csv
direc=os.path.splitext(os.path.splitext(finalfn)[0])[0]
anfn=datapath+direc+"\\"+direc+"*_0.trace"
anfn=glob.glob(anfn)
if len(anfn) <1:
#print "cannot find ",anfn,finalfn
#sys.exit()
return None,None,None,None
else:
anfn=anfn[0]
#print os.path.exists(anfn)
sb=statbreaks.brkrpt(anfn)
return sb.breaks(), sb.years(), sb.ys(), sb.segments()
def __init__(self, rootname):
self.__rootname = rootname
self.__finalfile = rootname + ".final.csv"
self.__lines = []
with open(self.__finalfile) as f:
lines = f.readlines()
for line in lines:
yrset = line[: line.find('"', 1) + 1]
tail = line[line.find('"', 1) + 2 : -1]
self.__lines.append([yrset, tail.split(" ")])
# print self.__lines[0][1][0] will be the frequency count
# now we have a list of lists [[str(breaks), [str(parameters)]]]
# find the first line with the max freq
maxi = 0
maxix = 0
maxs = []
self.__METADATA = []
for l in range(len(self.__lines)):
try:
fr = int(self.__lines[l][1][0])
if fr > maxi:
maxi = fr
maxix = l
maxs = self.__lines[l][0]
except: # metadate at bottom
self.__METADATA.append([rootname, self.__lines[l]])
pass
# open any trace file in the _self.__rootnamedir and extract the years and yi values
try:
samplefn = glob.glob(rootname + "\\*.trace")[0]
except:
samplefn = ""
sb = statbreaks.brkrpt(samplefn)
self.__years = sb.years()
self.__ys = sb.ys()
self.__maxset = maxs
# now finds all the lines nelonging to modal
self.__analysis = []
for line in self.__lines:
if maxs == line[0]:
self.__analysis.append(line[1])
def newstyle(csvrepfn,outfn): #from csv files
outf=open(outfn, "w")
with open(csvrepfn,'r') as f:
lines=f.readlines()
for line in lines:
#print line
if line[:5] == '"Seq"':
print >>outf,csvHeader+',":",'+cmip5_info.header+ancovaHeader
else:
if line.find("no breaks")>=0:
nobreaks = True
else:
nobreaks = False
words=line.split('"')
try:
# nobreaks=False
fn=os.path.basename(words[0].split(",")[1])
except Exception as e:
print "Oops", line, words, str(e)
if line.find("no breaks")>=0:
nobreaks = True
pass
else:
raise e
#print fn+":"
if len(words) < 2:
print "oops2" , line, words
if not nobreaks:
yset=eval(words[1])
#print "YSET", yset
yrss=words[2].split(',')[8:11]
yr=str(yrss[0])
pre=str(yrss[1])
post=str(yrss[2])
# print yr,pre,post
dfn=fn[:-len('final.csv')-1]
if dfn[:len("Global-temps")]=="Global-temps":
dfn = "Global temps"+dfn[len("Global-temps"):]
tfn=os.environ["HOMEPATH"]+"\\Documents\\abrupt\\c_test\\"+dfn
#print fn, tfn
if dfn[:len("Global")]=="Global" or dfn[0] == 'r':
print "do",tfn
try:
direc=glob.glob(os.environ["HOMEPATH"]+"\\Documents\\abrupt\\c_test\\"+dfn+'/*.trace')[0]
bp=statbreaks.brkrpt(direc)
#print bp.years()
yrix=int(list(bp.years()).index(float(yr)))
preix=int(list(bp.years()).index(float(pre)))
postix=int(list(bp.years()).index(float(post)))
#print bp.years()[preix:postix+1]
ancova=doancova.ANCOVA(bp.years()[preix:postix+1], bp.ys()[preix:postix+1], bp.years()[yrix],Rpath='"C:\\Program Files\\R\\R-3.0.2\\bin\\Rscript.exe"')
ancova.parseResult(ancova.results())
print >>outf,line[:-1],',":"',
if dfn[:len("Global")]=="Global":
info=cmip3_info.matchfn(tfn)
for inf in info:
print >>outf, ",",inf,
elif dfn[0] == 'r':
info=cmip5_info.matchfn(tfn)
for inf in info:
print >>outf, ",",inf,
print >>outf, ',":"',
for inf in [
ancova._ANCOVA__breakYear,
ancova.firstSlope(),
formatprobs(ancova.firstSlopeProb()),
ancova.firstRsq(),
ancova.secondSlope(),
formatprobs(ancova.secondSlopeProb()),
ancova.secondRsq(),
ancova.shift(),
formatprobs(ancova._ANCOVA__prDiffShift),
":",
formatprobs(ancova.changeOfTrendProb(), ancova._ANCOVA__prDiffShift),
formatprobs(ancova.changeOfTrendProb()* ancova._ANCOVA__prDiffShift)
]:
print >>outf,",",inf,
print >>outf
#if
except Exception as e:
fl=os.environ["HOMEPATH"]+"\\Documents\\abrupt\\c_test\\"+fn[:-len('final.csv')-1]+'/*.trace'
print >>outf,line[:-1],"Error getting files",fl
print "using ", fl
#print glob.glob(fl)
#raise Exception(str(e))
pass
#print direc,
outf.close()
print "done"
def __init__(self, fn): self.__filename = fn self.__data=statbreaks.brkrpt(fn) self.__yearsegs, self.__ysegs, self.__xsegs = self.__data.segments()
return self.__prDiffTrend
def firstRsq(self):
return self.__PriorRsq
def secondRsq(self):
return self.__PostRsq
def prBreak(self):
return self__prbreak
if __name__ == "__main__":
fn1=os.environ["HOMEPATH"]+"\\Documents\\abrupt\\c_test\\GISSTEMP_v3_April_2015_withDiffs_from_2014_24N44N\\GISSTEMP_v3_April_2015_withDiffs_from_2014.csv_24N44N_0.trace"
#os.environ["HOMEPATH"]+"\\Documents\\abrupt\\4Roger_Nature_SVN_264\\HadCRUT.4.2.0.0.annual_ns_avg//HadCRUT.4.2.0.0.annual_ns_avg.txt_0.trace")
bp=statbreaks.brkrpt(fn1)
yearsegs, ysegs, xsegs = bp.segments()
pre98Years=yearsegs[-2]
pre98temps=ysegs[-2]
post98Years=yearsegs[-1]
post98temps=ysegs[-1]
Years=list(pre98Years[:])
Years.extend(list(post98Years[:]))
ys=list(pre98temps[:])
ys.extend(list(post98temps[:]))
print bp.breaks()
ancova=ANCOVA(Years, ys, bp.breaks()[-2], Rpath='"C:\\Program Files\\R\\R-3.0.2\\bin\\Rscript.exe"')
print ancova.parseResult(ancova.results())
for l in ancova.formatted():
passyr=float(word[:-1].strip())
except:
passyr=word[:-2].strip()
pass
if state == 2:
#find the last date should be returned
words = line.split()
word=words[0]
while word[0] == "[" or word[-1]!= "]":
word=words.pop(0)
try:
if word[-2] == ".":
finalyr=float(word[:-2].strip())
else:
finalyr=float(word[:-1].strip())
except:
finalyr=word[:-2].strip()
pass
if state==3:
filename=line
state = 4
if state == 4:
if passyr!= finalyr:
sf=statbreaks.brkrpt(filename[:-1])
brks=sf.breaks()
if passyr > brks[-2]:
print passyr== finalyr,passyr, finalyr, brks[-2:], filename[:-1]
state=0