def toNum():
for size in ["all"]:
iFile = "%s/online_all.csv" % datPath
oFile1 = "%s/onlineT_all.csv" % datPath # data for Take.core
oFile2 = "%s/onlineO_all.basket" % datPath # data for Orange
oFile3 = "%s/onlineM_all.csv" % datPath # data for Take
f = None
f <<= nm.mcut(f="InvoiceNo,StockCode", i=iFile)
f <<= nm.muniq(k="InvoiceNo,StockCode")
f <<= nm.mfldname(q=True, o=oFile3)
f.run(msg=debug)
st = None
st <<= nm.mcut(f="StockCode", i=iFile)
st <<= nm.muniq(k="StockCode")
st <<= nm.mnumber(s="StockCode", a="num")
f = None
f <<= nm.mjoin(k="StockCode", m=st, f="num", i=iFile)
f <<= nm.mcut(f="InvoiceNo,num:StockCode")
f <<= nm.mtra(k="InvoiceNo", f="StockCode")
f <<= nm.mcut(f="StockCode", nfno=True, o=oFile1)
f.run(msg=debug)
os.system("tr ' ' ',' <%s >%s" % (oFile1, oFile2))
def __init__(self,db): self.db=db self.eArgs=None self.type =None self.minCnt=None self.minSup=None self.maxCnt=None self.maxSup=None self.minLen=None self.maxLen=None self.top =None self.skipTP=False #self.size =None self.pFile =None self.tFile =None self.msgoff = True self.temp=mtemp.Mtemp() self.db = db # 入力データベース self.file=self.temp.file() items=self.db.items # アイテムをシンボルから番号に変換する。 f = nm.mjoin(k=self.db.itemFN,K=items.itemFN,m=items.file,f=items.idFN,i=self.db.file) f <<= nm.mcut(f=self.db.idFN+","+items.idFN) f <<= nm.mtra(k=self.db.idFN,f=items.idFN) f <<= nm.mcut(f=items.idFN,nfno=True,o=self.file) f.run()
def __mkMap(key, nf, ni, ef1, ef2, ei, oFile):
# leaf nodeの構築
infL = [
nm.mcommon(k=ef1, K=key, m=ei, r=True, i=ei).mcut(f="%s:nam" % (ef1)),
nm.mcommon(k=ef2, K=key, m=ei, r=True, i=ei).mcut(f="%s:nam" % (ef2))
]
if ni:
infL.append(
nm.mcommon(k=nf, K=key, m=ei, r=True,
i=ni).mcut(f="%s:nam" % (nf)))
xleaf = nm.muniq(i=infL, k="nam")
xleaf <<= nm.msetstr(v=1, a="leaf")
if ni:
inp = [
nm.mcut(f="%s:nam" % (nf), i=ni),
nm.mcut(f="%s:nam" % (key), i=ni)
]
else:
inp = [
nm.mcut(f="%s:nam" % (ef1), i=ei),
nm.mcut(f="%s:nam" % (ef2), i=ei),
nm.mcut(f="%s:nam" % (key), i=ei)
]
f = None
f <<= nm.muniq(k="nam", i=inp)
f <<= nm.mjoin(k="nam", m=xleaf, f="leaf", n=True)
# nullは最初に来るはずなので、mcalでなくmnumberでもnullを0に採番できるはずだが念のために
f <<= nm.mcal(c='if(isnull($s{nam}),0,line()+1)', a="num")
f <<= nm.mnullto(f="nam", v="##NULL##", o=oFile)
f.run()
def pair2tra(self, ei, ef1, ef2, traFile, mapFile1, mapFile2):
f1 = None
f1 <<= nm.mcut(f="{}:node1".format(self.ef1), i=self.ei)
f1 <<= nm.mdelnull(f="node1")
f1 <<= nm.muniq(k="node1")
f1 <<= nm.mnumber(s="node1", a="num1", o=mapFile1)
f2 = None
f2 <<= nm.mcut(f="{}:node2".format(self.ef2), i=self.ei)
f2 <<= nm.mdelnull(f="node2")
f2 <<= nm.muniq(k="node2")
f2 <<= nm.mnumber(s="node2", a="num2", o=mapFile2)
f3 = None
f3 <<= nm.mcut(f="{}:node1,{}:node2".format(self.ef1, self.ef2),
i=self.ei)
f3 <<= nm.mjoin(k="node1", m=f1, f="num1")
f3 <<= nm.mjoin(k="node2", m=f2, f="num2")
f3 <<= nm.mcut(f="num1,num2")
f3 <<= nm.msortf(f="num1,num2%n")
f3 <<= nm.mtra(k="num1", s="num2%n", f="num2")
f3 <<= nm.msortf(f="num1%n")
f3 <<= nm.mcut(f="num2", nfno=True, o=traFile)
f3.run()
os.system("cat " + traFile)
def edge2mtx(self, ei, itra, map1, map2):
p1 = nm.mcut(f=self.ef1, i=ei)
p1 <<= nm.muniq(k=self.ef1)
p1 <<= nm.mdelnull(f=self.ef1)
p1 <<= nm.mnumber(q=True, a="num1", S=1, o=map1)
p2 = nm.mcut(f=self.ef2, i=ei)
p2 <<= nm.muniq(k=self.ef2)
p2 <<= nm.mdelnull(f=self.ef2)
p2 <<= nm.mnumber(q=True, a="num2", S=1, o=map2)
runp = None
runp <<= nm.mcut(f=[self.ef1, self.ef2], i=ei)
runp <<= nm.mjoin(k=self.ef1, m=p1, f="num1")
runp <<= nm.mjoin(k=self.ef2, m=p2, f="num2")
runp <<= nm.mcut(f="num1,num2")
runp <<= nm.mtra(k="num1", f="num2")
runp <<= nm.msortf(f="num1%n")
runp <<= nm.mcut(f="num2", nfno=True)
runp <<= nm.cmd("tr ',' ' '")
runp <<= nm.mwrite(o=itra)
#runp <<= nm.mcut(f="num2",nfno=True,o=wff1)
runp.run()
def __init__(self, db, outtf=True):
self.size = None
self.pFile = None
self.tFile = None
self.temp = mtemp.Mtemp()
self.db = db # 入力データベース
self.file = self.temp.file()
self.outtf = outtf
self.weightFile = {}
self.posWeight = {}
self.sigma = {}
self.msgoff = True
items = self.db.items
for cName, posSize in db.clsNameRecSize.items():
self.weightFile[cName] = self.temp.file()
self.posWeight[cName] = self.calOmega(posSize)
f = nm.mcut(nfno=True, f=self.db.clsFN, i=self.db.cFile)
f <<= nm.mchgstr(nfn=True,
f=0,
O=-1,
o=self.weightFile[cName],
c="%s:%s" % (cName, self.posWeight[cName]))
f.run()
# アイテムをシンボルから番号に変換する。
f = nm.mjoin(k=self.db.itemFN,
K=items.itemFN,
i=self.db.file,
m=items.file,
f=items.idFN)
f <<= nm.mcut(f=self.db.idFN + "," + items.idFN)
f <<= nm.mtra(k=self.db.idFN, f=items.idFN)
f <<= nm.mcut(f=items.idFN, nfno=True, o=self.file)
f.run()
def mkIndex(oFile, iFile):
f = None
f <<= nm.mcat(i=iFile)
f <<= nm.mcut(f="date,c")
f <<= nm.mavg(k="date", f="c")
f <<= nm.mcal(c="round(${c},1)", a="i")
f <<= nm.mcut(f="date,i", o="%s" % oFile)
f.run()
def mcut(iFile, loop):
sec = []
for i in range(loop):
st = time.time()
nm.mcut(f="id,key1,key2,key3,int1,int2,float1,float2,date,time",
i=iFile,
o="oFile").run()
sec.append(time.time() - st)
return sec
def __mkEdge(key, ef1, ef2, el, ec, ed, ev, ei, norm, mapFile, oFile):
# mcal cat用のlabel項目の作成
label = []
if el:
for nml in el:
label.append("$s{" + nml + "}")
evcdStr = []
#エッジの拡大率は固定
er = 10
if ev:
evcdStr.append(ev + ":ev")
if ec:
evcdStr.append(ec + ":ec")
if ed:
evcdStr.append(ed + ":ed")
f = None
if el:
f <<= nm.mcal(c='cat(\"_\",%s)' % (','.join(label)), a="##label", i=ei)
else:
f <<= nm.msetstr(v="", a="##label", i=ei)
if len(evcdStr) == 0:
f <<= nm.mcut(f="%s:key,%s:nam1,%s:nam2,##label:el" % (key, ef1, ef2))
else:
f <<= nm.mcut(f="%s:key,%s:nam1,%s:nam2,##label:el,%s" %
(key, ef1, ef2, ','.join(evcdStr)))
if not ev:
f <<= nm.msetstr(v="", a="ev")
if not ed:
f <<= nm.msetstr(v="", a="ed")
if not ec:
f <<= nm.msetstr(v="", a="ec")
f <<= nm.mnullto(f="key", v="##NULL##")
f <<= nm.mjoin(k="key", K="nam", m=mapFile, f="num:keyNum")
f <<= nm.mjoin(k="nam1", K="nam", m=mapFile, f="num:num1,leaf:leaf1")
f <<= nm.mjoin(k="nam2", K="nam", m=mapFile, f="num:num2,leaf:leaf2")
if norm:
f <<= nm.mnormalize(f="ev:ev2", c="range")
f <<= nm.mcal(c='${ev2}*(%s-1)+1' % (er), a="evv")
f <<= nm.mcut(
f="key,nam1,nam2,keyNum,num1,num2,el,evv:ev,ed,ec,leaf1,leaf2",
o=oFile)
else:
f <<= nm.mcut(
f="key,nam1,nam2,keyNum,num1,num2,el,ev,ed,ec,leaf1,leaf2",
o=oFile)
f.run()
def __readNodeCSV(self, nodeFile=None, title=None):
if nodeFile is not None:
self.nFileN = nodeFile
if title is None:
mcmd = nm.mcut(nfni=True, f="0:n", i=self.nFileN)
else:
self.nodeFN = title
mcmd = nm.mcut(f="%s:n" % (self.nodeFN), i=self.nFileN)
self.__wfN = self.__tempW.file()
mcmd <<= nm.mnumber(q=True, a="no", o=self.__wfN)
mcmd.run()
def __readEdgeCSV(self, edgeFile=None, title1=None, title2=None):
self.eFileN = edgeFile
if title1 is None or title2 is None:
mcmd = nm.mcut(nfni=True, f="0:e1,1:e2", i=self.eFileN)
else:
self.edgeFN1 = title1
self.edgeFN2 = title2
mcmd = nm.mcut(f="%s:e1,%s:e2" % (self.edgeFN1, self.edgeFN2),
i=self.eFileN)
self.__wfE = self.__tempW.file()
mcmd <<= nm.mnumber(q=True, a="no", o=self.__wfE)
mcmd.run()
def convSim(self, ifile, ofile, map1, logDir):
f = None
f <<= nm.mcut(nfni=True, f="0:tra", i=ifile)
f <<= nm.msed(f="tra", c=' $', v="")
f <<= nm.mnumber(q=True, S=1, a="num1")
f <<= nm.mtra(r=True, f="tra:num11")
f <<= nm.mnumber(q=True, S=1, a="order")
f <<= nm.mcal(c='${num11}+1', a="num1")
f <<= nm.mjoin(k="num1", m=map1, f=self.ef1)
f <<= nm.mtra(k="num0", s="order%n,num1%n", f=self.ef1)
f <<= nm.mcut(f=self.ef1, o="{}/{}".format(logDir, ofile))
f.run()
def calGsize(self, file):
edgesize = nu.mrecount(i=file, nfni=True)
f = None
f <<= nm.mcut(i=file, f="0:tra", nfni=True)
f <<= nm.mtra(f="tra")
f <<= nm.muniq(k="tra")
f <<= nm.mcount(a="cnt")
f <<= nm.mcut(f="cnt")
f <<= nm.writelist(dtype="cnt:int")
nodesize = f.run()[0][0]
return nodesize, edgesize
def readCSV(iParams):
iFile = iParams["iFile"]
sidF = iParams["sid"]
eidF = iParams["time"]
itemF = iParams["item"]
temp = Mtemp()
xxdatPath = temp.file()
mkDir(xxdatPath)
# classファイルの処理
if "cFile" in iParams:
cFile = iParams["cFile"]
csidF = iParams["csid"]
classF = iParams["cNames"]
f = None
f <<= nm.mcut(f="%s:sid,%s:eid,%s:item" % (sidF, eidF, itemF), i=iFile)
f <<= nm.mdelnull(f="sid,eid,item")
f <<= nm.muniq(k="sid,eid,item")
f <<= nm.mjoin(k="sid", K=csidF, m=cFile, f="%s:class" % (classF))
f <<= nm.msep(s="sid,eid%n,item",
d="%s/${class}" % (xxdatPath),
p=True)
f.run()
classNames = glob.glob("%s/*" % (xxdatPath))
classNames = [os.path.basename(path) for path in classNames]
else:
f = None
f <<= nm.mcut(f="%s:sid,%s:eid,%s:item" % (sidF, eidF, itemF), i=iFile)
f <<= nm.mdelnull(f="sid,eid,item")
f <<= nm.muniq(k="sid,eid,item")
f <<= nm.msortf(f="sid,eid%n,item", o="%s/single" % (xxdatPath))
f.run()
classNames = ["single"]
datas = {}
for name in classNames:
dataFile = "%s/%s" % (xxdatPath, name)
datas[name] = _readCSV_sub(dataFile)
return datas
def nm2(iFile):
f = None
f <<= nm.mcut(f="id,date,c", i=iFile)
f <<= nm.mwindow(k="id", wk="date:win", t=3)
f <<= nm.mavg(k="id,win", f="c")
f <<= nm.writelist(dtype="win:int,date:int,c:float")
r = f.run()
def convOrg(self,xxmap,xxout,ofile): xx1 = nm.mnumber(q=True,S=0,a="num",i=xxout) f = nm.mjoin(k="num",f="cls",m=xx1,i=xxmap) f <<= nm.mcut(f="node,cls", o=ofile) f.run()
def conv(self,iFile,idFN,itemFN,oFile,mapFile): temp=nu.Mtemp() xxtra=temp.file() # 入力ファイルのidがnilの場合は連番を生成して新たなid項目を作成する。 f0 = nm.mcut(f=itemFN+":##item",i=iFile) f0 <<= nm.mcount(k="##item",a="##freq") f0 <<= nm.mnumber(s="##freq%nr",a="##num",o=mapFile) f0.run() f1 = nm.mjoin(k=itemFN,K="##item",f="##num",m=mapFile,i=iFile,o=xxtra) f1.run() #f1 <<= nm.mtra(k=idFN,f="##num") #f1 <<= nm.mnumber(q=True,a="##traID") #f1 <<= nm.mcut(f="##num",nfno=True,o=oFile) extTake.mmaketra(i=xxtra,o=oFile,f="##num",k=idFN) #f1.run() # ##item,##freq%0nr,##num # b,4,0 # d,4,1 size=nu.mrecount(i=oFile,nfn=True) return size
def addTaxo(self, taxonomy):
self.taxonomy = taxonomy
self.items.addTaxo(self.taxonomy) # アイテムクラスにtaxonomyを追加する
tFile = taxonomy.file
itemFN = taxonomy.itemFN
taxoFN = taxonomy.taxoFN
tf = nu.Mtemp()
xx1 = tf.file()
cpara = "%s,%s:%s" % (self.idFN, taxoFN, self.itemFN)
f = None
f <<= nm.mnjoin(i=self.file,
k=self.itemFN,
K=itemFN,
f=taxoFN,
m=tFile)
f <<= nm.mcut(f=cpara, o=xx1)
f.run()
ipara = "%s,%s" % (self.file, xx1)
kpara = "%s,%s" % (self.idFN, self.itemFN)
xx2 = tf.file()
nm.mcat(i=ipara).muniq(k=kpara, o=xx2).run(msg="on")
self.file = self.temp.file()
shutil.move(xx2, self.file)
def __readEdgeList(self, edgeList=None, title1=None, title2=None):
if title1 is not None and title2 is not None:
self.edgeFN1 = title1
self.edgeFN2 = title2
mcmd = nm.mcut(nfni=True, f="0:e1,1:e2", i=edgeList)
self.__wfE = self.__tempW.file()
mcmd <<= nm.mnumber(q=True, a="no", o=self.__wfE)
mcmd.run()
def runmain(self,edgeFile):
import re
baseName = re.sub('\.edge$',"",edgeFile)
name = re.sub('^.*\/',"",baseName)
if self.ni :
nodeFile=self.ni
else :
nodeFile=re.sub('\.edge$',".node",edgeFile)
# convert the original graph to one igraph can handle
temp=Mtemp()
xxnum = temp.file()
xxmap = temp.file()
xxout = temp.file()
xxscp = temp.file()
xxweight=temp.file()
nodeSize=self.g2pair(
nodeFile,self.nf,edgeFile,self.ef1,self.ef2,self.ew,xxnum,xxmap,xxweight
)
# generate R script, and run
self.genRscript(
self.directed,self.norm,self.mode,
xxnum, xxweight, self.ew,nodeSize, xxout, xxscp
)
if self.verbose :
os.system("R --vanilla -q < %s"%(xxscp))
else:
os.system("R --vanilla -q --slave < %s 2>/dev/null "%(xxscp))
f = None
f <<= nm.mnumber(q=True,S=0,a="num",i=xxout)
f <<= nm.mjoin(k="num",f="node",m=xxmap )
outf = self.oPath + "/" +name + ".csv"
if self.nf :
f <<= nm.mcut(f="node:%s,degree,cc,components,betweenness,closeness,page_rank"%(self.nf),o=outf)
else:
f <<= nm.mcut(f="node,degree,cc,components,betweenness,closeness,page_rank",o=outf)
f.run()
def g2pair(self,ni,nf,ei,ef1,ef2,ew,numFile,mapFile,weightFile):
inobj = []
inobj.append(nm.mcut(f="%s:node"%(ef1),i=ei ).msetstr(a="flag",v=0))
inobj.append(nm.mcut(f="%s:node"%(ef2),i=ei ).msetstr(a="flag",v=0))
if nf :
inobj.append(nm.mcut(f="%s:node"%(nf),i=ni ).msetstr(a="flag",v=1))
f = nm.mbest(i=inobj,k="node",s="flag",fr=0,size=1 )
# isolated nodes are set to the end of position in mapping file.
# S= must start from 0 (but inside R vertex number will be added one)
f <<= nm.mnumber(s="flag,node",a="num",S=0,o=mapFile)
f.run()
f = None
f <<= nm.mcut(f=[ef1,ef2] , i=ei)
f <<= nm.mjoin( k=ef1 , K="node" , m=mapFile ,f="num:num1")
f <<= nm.mjoin( k=ef2 , K="node" , m=mapFile ,f="num:num2")
f <<= nm.mcut(f="num1,num2")
f <<= nm.mfsort(f="num1,num2")
f <<= nm.msortf(f="num1%n,num2%n",nfno=True)
f <<= nm.cmd("tr ',' ' ' " )
f <<= nm.mwrite(o=numFile)
f.run()
nodeSize=mrecount(i=mapFile)
if ew:
nm.mcut(f=ew,i=ei,o=weightFile).run()
else:
ew="weight"
nm.msetstr(v=1,a=ew,i=ei).mcut(f=ew,o=weightFile).run()
return nodeSize
def mnest2tree(ei, ef, k, ni=None, nf=None, ev=None, no=None, eo=None):
# paracheck追加
efs = ef.split(",")
ef1 = efs[0]
ef2 = efs[1]
f = nm.mcut(f="%s:#orgKey,%s:#orgEf1,%s:#orgEf2" % (k, ef1, ef2), i=ei)
temp = mtemp.Mtemp()
of = temp.file()
with _nu.mcsvout(o=of, f="#orgKey,#orgEf1,#orgEf2,#ef1,#ef2") as oCSV:
for flds in f:
orgKey = flds[0]
orgEf1 = flds[1]
orgEf2 = flds[2]
oCSV.write([orgKey, orgEf1, orgEf2, orgKey, orgEf1])
oCSV.write([orgKey, orgEf1, orgEf2, orgKey, orgEf2])
f = None
f <<= nm.mjoin(k="#orgKey,#orgEf1,#orgEf2", K=[k, ef1, ef2], m=ei,
i=of) # 全項目join
if ev:
f <<= nm.mavg(k="#ef1,#ef2", f=ev)
else:
f <<= nm.muniq(k="#ef1,#ef2")
f <<= nm.mcut(r=True, f="#orgKey,#orgEf1,#orgEf2")
f <<= nm.mfldname(f="#ef1:%s,#ef2:%s" % (ef1, ef2), o=eo)
f.run()
if ni:
head = nu.mheader(i=ni)
fldnames = [s for s in head if s != nf]
commas = ',' * (len(fldnames) - 1)
f0 = None
f0 <<= nm.mcut(f="%s:%s" % (ef1, nf), i=eo)
f0 <<= nm.muniq(k=nf)
f0 <<= nm.mcommon(k=nf, m=ni, r=True)
f0 <<= nm.msetstr(v=commas, a=fldnames)
f = nm.mcut(f=k, r=True, i=[ni, f0])
f <<= nm.msetstr(v="", a=k, o=no)
f.run()
def calTime(iFile, oFile):
f = None
f <<= nm.mnumber(q=True, a="id", i=iFile)
f <<= nm.mcal(c='$s{mean}+"("+$s{sd}+")"', a="time")
f <<= nm.m2cross(k="method", s="dataSize", f="time")
f <<= nm.msortf(f="id%n")
f <<= nm.mcut(f="method,10000:small,1000000:middle,100000000:large")
f <<= nm.mfldname(q=True, o=oFile)
f.run()
def replaceFile(self, train):
self.file = train
# レコード数の計算
self.recCnt = nu.mrecount(i=self.file)
xx1 = nm.mcut(f=self.idFN, i=self.file).muniq(k=self.idFN).mcount(
a='__cnt').mcut(a='__cnt').run()
self.traSize = int(xx1[0][0])
def __init__(self, db, outtf=True):
self.size = None
self.temp = mtemp.Mtemp()
self.db = db # 入力データベース
self.file = self.temp.file()
items = self.db.items
self.outtf = outtf
self.top = None
self.msgoff = True
# アイテムをシンボルから番号に変換する。
f = nm.mjoin(k=self.db.itemFN,
K=items.itemFN,
m=items.file,
f=items.idFN,
i=self.db.file)
f <<= nm.mcut(f=self.db.idFN + "," + self.db.timeFN + "," + items.idFN)
f <<= nm.mtra(k=self.db.idFN, s=self.db.timeFN + "%n", f=items.idFN)
f <<= nm.mcut(f=items.idFN, nfno=True, o=self.file)
f.run()
def tra2tbl(iFile,
oFile,
tidFld,
itemFld,
null=0,
dummy=True,
aggFld=None,
aggStat=None,
klassFld=None):
klass = None
if klassFld is not None and klassFld != "":
klass <<= nm.mcut(f=tidFld + "," + klassFld, i=iFile)
klass <<= nm.muniq(k=tidFld)
f = None
# aggFldが指定されていればセル項目を集計
if aggFld is not None and aggFld != "":
f <<= nm.mcut(f=tidFld + "," + itemFld + "," + aggFld, i=iFile)
f <<= nm.mstats(k=tidFld, f="%s:_cell" % (aggFld), c=aggStat)
# aggFldが指定されていなければカウント
else:
f <<= nm.mcut(f=tidFld + "," + itemFld, i=iFile)
# アイテムが出現したかどうか
if dummy:
f <<= nm.muniq(k=tidFld + "," + itemFld)
f <<= nm.msetstr(v=1, a="_cell")
# アイテムが何件出現したかどうか
else:
f <<= nm.mcount(k=tidFld + "," + itemFld, a="_cell")
# 横展開
f <<= nm.m2cross(k=tidFld, s=itemFld, f="_cell")
# クラス項目が指定されていれば結合する
if klassFld is not None and klassFld != "":
f <<= nm.mjoin(k=tidFld, m=klass, f=klassFld)
# null値を一斉に置換する
f <<= nm.mnullto(f="*", v=null, o=oFile)
f.run(msg="on")
def __init__(self,iFile,itemFN,taxoFN): # アイテムの項目名(=>String) self.itemFN = None # 分類の項目名(=>String) self.taxoFN = None # アイテムの種類数(=>Fixnum) self.itemSize = None # 分類の種類数(=>Fixnum) self.taxoSize = None # taxonomyデータファイル名(=>String) self.file = None self.temp = mtemp.Mtemp() self.iFile = iFile self.iPath = os.path.abspath(self.iFile) self.itemFN = itemFN self.taxoFN = taxoFN # item順に並べ替えてpathに書き出す self.file = self.temp.file() para_it = self.itemFN +"," + self.taxoFN nm.mcut(f=para_it,i=self.iFile).muniq(k=para_it,o=self.file).run(msg="on") f = nm.mcut(f=self.itemFN,i=self.iFile) f <<= nm.mtrafld(f=self.itemFN,a="__fld",valOnly=True) f <<= nm.mtra(f="__fld",r=True) f <<= nm.muniq(k="__fld") f <<= nm.mcount(a="size") f <<= nm.mcut(f="size") xx1 = f.run() self.itemSize = int(xx1[0][0]) xx2 = nm.mcut(f=self.taxoFN+":item",i=self.file).muniq(k="item").mcount(a="size").mcut(f="size").run() self.taxoSize = int(xx2[0][0])
def stats_h(self,b): field=self.field_w.getValue() self.parent.msg_w.value="\"%s\" 項目の統計量を計算中..."%(field) f=None f<<=nm.mcut(f=field,i=self.iName_w.value) f<<=nm.msummary(f=field,c="min,median,mean,max") f<<=nm.writelist(header=True) rsl=f.run() #print(rsl) self.result_w.value=str(rsl) self.parent.msg_w.value="\"%s\" 項目の統計量を計算中...完了"%(field)
def convRsl(self, ifile, ofile, map1, map2, logDir=None):
# 上記iterationで収束したマイクロクラスタグラフを元の節点文字列に直して出力する
#MCMD::msgLog("converting the numbered nodes into original name ...")
f = None
f <<= nm.mcut(nfni=True, f="0:tra", i=ifile)
f <<= nm.msed(f="tra", c=' $', v="")
f <<= nm.mnumber(q=True, S=1, a="num1")
f <<= nm.mtra(r=True, f="tra:num2")
f <<= nm.mjoin(k="num2", m=map2, f=self.ef2)
f <<= nm.mjoin(k="num1", m=map1, f=self.ef1)
f <<= nm.msortf(f="num1%n,num2%n")
f <<= nm.mcut(f=[self.ef1, self.ef2])
if logDir:
f <<= nm.mfldname(q=True, o="{}/#{ofile}".format(logDir, ofile))
else:
f <<= nm.mfldname(q=True, o=ofile)
f.run()
def __readNodeList(self, nodeList=None, title=None):
if nodeList is not None:
if title is not None:
self.nodeFN = title
self.__wfN = self.__tempW.file()
mcmd = nm.m2cross(nfn=True,
a="num,node",
f="0-%d" % (len(nodeList) - 1),
i=[nodeList])
mcmd <<= nm.mcut(nfni=True, f="1:n,0:no", o=self.__wfN)
mcmd.run()