def exe(self, script_w):
params = {}
params["iFile"] = self.iName_w.value
params["mFile"] = self.mName_w.value
params["oPath"] = self.oPath_w.value
params["oFile"] = self.oFile_w.value
if not wlib.iFileCheck(params["iFile"], self.parent.msg_w):
return False
if not wlib.iFileCheck(params["mFile"], self.parent.msg_w):
return False
if not wlib.blankCheck(params["oFile"], "出力ファイル名", self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"], self.parent.msg_w):
return False
iKey = self.iKey_w.getValue()
mKey = self.mKey_w.getValue()
field = self.field_w.getValue()
iOuter = self.iOuter_w.value
mOuter = self.mOuter_w.value
if not wlib.blankCheck(field, "結合する項目", self.parent.msg_w):
return False
par = []
par.append("k='%s'" % (iKey))
if mKey != "":
par.append("K='%s'" % (mKey))
par.append("m='%s'" % (params["mFile"]))
par.append("f='%s'" % (field))
if iOuter:
par.append("n=True")
if mOuter:
par.append("N=True")
par.append("i='%s'" % (params["iFile"]))
par.append("o='%s/%s'" % (params["oPath"], params["oFile"]))
header = """
#################################
# mjoin_w.pyの自動生成スクリプト
# version: %s
# 実行日時: %s
#################################
""" % (self.version, self.date)
lib = """
import nysol.mcmd as nm
nm.setMsgFlg(True)
print("#### START")
"""
script = """
f=None
f<<=nm.mjoin(%s)
f.run(msg='on')
print("#### END")
""" % (",".join(par))
script_w.value = header + lib + script
return True
def exe(self, script_w):
params = {}
params["iFile"] = self.fName_w.value
params["oPath"] = self.oPath_w.value + "/" + self.oDir_w.value
if not wlib.iFileCheck(params["iFile"], self.parent.msg_w):
return False
if not wlib.blankCheck(self.oDir_w.value, "出力dir名", self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"], self.parent.msg_w):
return False
deepOutput = self.parent.deepOutput_w.value
params["id_"] = self.id_w.getValue()
params["y"] = self.y_w.getValue()
params["nums"] = list(self.num_w.getValue())
params["cats"] = list(self.cat_w.getValue())
params["minSamplesLeaf"] = self.minSamplesLeaf_w.value
if params["y"] in params["nums"] or params["y"] in params["cats"]:
self.parent.msg_w.value = "##ERROR: 説明変数に目的変数が含まれています:%s" % params[
"y"]
return False
script1 = wlib.readSource("nysol.mining.ctree", "ctree", deepOutput)
script2 = wlib.readSource("nysol.mining.csv2df")
script3 = """
print("#### START")
# 出力ディレクトリを作成する
os.makedirs("{oPath}",exist_ok=True)
# pandasデータを作成する
df,ds=csv2df("{iFile}", "{id_}", {nums}, ["{y}"], {cats})
xNames=ds.columns.to_list()
xNames.remove("{y}")
y=pd.DataFrame(ds.loc[:,"{y}"])
x=ds.loc[:,xNames]
config=dict()
config["max_depth"]=10
config["min_samples_leaf"]={minSamplesLeaf}
model=ctree(x,y,config)
model.build()
print("cv_minFunc",model.cv_minFun)
print("cv_minX",model.cv_minX)
print("score",model.score)
model.visualize()
model.save("{oPath}/model")
pred=model.predict(x)
pred.evaluate(y)
pred.save("{oPath}/pred")
print("#### END")
""".format(**params)
script_w.value = script1 + script2 + script3
return True
def exe(self, script_w):
params = {}
params["iFile"] = self.iName_w.value
params["oPath"] = self.oPath_w.value + "/" + self.oDir_w.value
if not wlib.iFileCheck(params["iFile"], self.parent.msg_w):
return False
if not wlib.blankCheck(self.oDir_w.value, "出力dir名", self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"], self.parent.msg_w):
return False
params["minFreq"] = self.minFreq_w.value
script1 = wlib.readSource("nysol.widget.songs.lib.features")
script2 = """
import os
print("#### START")
oPath="{oPath}"
os.makedirs(oPath,exist_ok=True)
features("{iFile}",oPath,{minFreq})
print("#### END")
""".format(**params)
# script tabにセット
script_w.value = script1 + script2
return True
def exe(self, script_w):
params = {}
params["iFile"] = self.iName_w.value
params["oPath"] = self.oPath_w.value
params["oFile"] = self.oFile_w.value
if not wlib.iFileCheck(params["iFile"], self.parent.msg_w):
return False
if not wlib.blankCheck(params["oFile"], "出力ファイル名", self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"], self.parent.msg_w):
return False
script1 = wlib.readSource("nysol.widget.songs.lib.parsing", "parsing")
script2 = """
import os
print("#### START")
oPath="{oPath}"
os.makedirs(oPath,exist_ok=True)
parsing_count("{iFile}","%s/{oFile}"%(oPath))
print("#### END")
""".format(**params)
# script tabにセット
script_w.value = script1 + script2
return True
def exe(self, script_w):
params = {}
params["iFile"] = self.iName_w.value
params["oPath"] = self.oPath_w.value
params["oFile"] = self.oFile_w.value
if not wlib.iFileCheck(params["iFile"], self.parent.msg_w):
return False
if not wlib.blankCheck(params["oFile"], "出力ファイル名", self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"], self.parent.msg_w):
return False
key = self.key_w.getValue()
field = self.field_w.getValue()
stat = self.stat_w.value
newfld = self.newfld_w.value
flds = []
if key != "":
flds.append(key)
flds.append(field)
params1 = "f='%s',i='%s'" % (",".join(flds), params["iFile"])
params2 = ""
if key != "":
params2 += "k='%s'" % (key)
if newfld == "":
params2 += ",f='%s'" % (field)
else:
params2 += ",f='%s:%s'" % (field, newfld)
params2 += ",c='%s',o='%s/%s'" % (stat, params["oPath"],
params["oFile"])
header = """
#################################
# mstats_w.pyの自動生成スクリプト
# version: %s
# 実行日時: %s
#################################
""" % (self.version, self.date)
lib = """
import nysol.mcmd as nm
nm.setMsgFlg(True)
print("#### START")
"""
script = """
f=None
f<<=nm.mcut(%s)
f<<=nm.mstats(%s)
f.run(msg='on')
print("#### END")
""" % (params1, params2)
script_w.value = header + lib + script
return True
def exe(self, script_w):
params = {}
params["iFile"] = self.iName_w.value
params["oPath"] = self.oPath_w.value
params["oFile"] = self.oFile_w.value
if not wlib.iFileCheck(params["iFile"], self.parent.msg_w):
return False
if not wlib.blankCheck(params["oFile"], "出力ファイル名", self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"], self.parent.msg_w):
return False
params["tid"] = self.tid_w.getValue()
params["item"] = self.item_w.getValue()
params["agg"] = self.agg_w.getValue()
params["klass"] = self.klass_w.getValue()
params["null"] = self.null_w.value
params["dummy"] = self.dummy_w.value
params["stat"] = self.stat_w.value
script1 = wlib.readSource("nysol.mining.tra2tbl")
script2 = """
print("#### START")
iFile="{iFile}"
oFile="{oPath}/{oFile}"
tidFld="{tid}"
itemFld="{item}"
null={null}
dummy={dummy}
aggFld="{agg}"
aggStat="{stat}"
klassFld="{klass}"
tra2tbl(iFile,oFile,tidFld,itemFld,null=null,dummy=dummy,aggFld=aggFld,aggStat=aggStat,klassFld=klassFld)
print("#### END")
""".format(**params)
script_w.value = script1 + script2
return True
def exe(self, script_w):
params = {}
params["iFile"] = self.iName_w.value
params["oPath"] = self.oPath_w.value
params["oFile"] = self.oFile_w.value
if not wlib.iFileCheck(params["iFile"], self.parent.msg_w):
return False
if not wlib.blankCheck(params["oFile"], "出力ファイル名", self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"], self.parent.msg_w):
return False
_head = self.head_w.value
_tail = self.tail_w.value
reverse = self.reverse_w.value
key = self.key_w.getValue()
field = self.field_w.getValue()
value = self.value_w.value
if value == "":
self.parent.msg_w.value = "##ERROR: 「値」が入力されていません。"
return False
if _head and _tail:
sub = False
head = False
tail = False
elif _head:
sub = False
head = True
tail = False
elif _tail:
sub = False
head = False
tail = True
else:
sub = True
head = False
tail = False
par = []
if key != "":
par.append("k=\"" + key + "\"")
par.append("f=\"" + field + "\"")
par.append("v=\"" + value + "\"")
if reverse:
par.append("r=True")
if head:
par.append("head=True")
if tail:
par.append("tail=True")
if sub:
par.append("sub=True")
par.append("i=\"" + params["iFile"] + "\"")
par.append("o=\"" + params["oPath"] + "/" + params["oFile"] + "\"")
header = """
#################################
# mselstr_w.pyの自動生成スクリプト
# version: %s
# 実行日時: %s
#################################
""" % (self.version, self.date)
lib = """
import nysol.mcmd as nm
nm.setMsgFlg(True)
print("#### START")
"""
script = """
f=None
f<<=nm.mselstr(%s)
f.run(msg='on')
print("#### END")
""" % (",".join(par))
script_w.value = header + lib + script
return True
def exe(self,script_w):
params={}
params["version"] = self.version
params["date"] = self.date
params["iFile"] = self.iName_w.value
params["oPath"] = self.oPath_w.value
params["oDir"] = self.oDir_w.value
if not wlib.iFileCheck(params["iFile"],self.parent.msg_w):
return False
if not wlib.blankCheck(params["oDir"],"出力dir名",self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"],self.parent.msg_w):
return False
params["tid"] = self.traID_w.getValue()
params["item"] = self.item_w.getValue()
params["minSup"] = self.minSup_w.value
params["sim"] = self.edgeType_w.value
params["nodeSup"] = self.nodeSup_w.value
if params["sim"]=="minConf":
params["th"] = self.minConf_w.value
params["simName"] = "C"
params["simFld"] = "confidence"
params["undirect"] = False
else:
params["th"] = self.minPMI_w.value
params["simName"] = "P"
params["simFld"] = "PMI"
params["undirect"] = True
if params["tid"]==params["item"]:
self.parent.msg_w.value="##ERROR: トランザクションIDとアイテム項目が同じ項目になってます"
return False
script="""
import os
import nysol.take as nt
import nysol.mcmd as nm
import nysol.view.mnetpie as nnpie
nm.setMsgFlg(True)
print("#### START")
# 出力ディレクトリを作成する
op="{oPath}"+"/"+"{oDir}"
os.makedirs(op,exist_ok=True)
nt.mtra2gc(i="{iFile}",
tid="{tid}",
item="{item}",
no=op+"/node.csv",
eo=op+"/edge.csv",
s={minSup},
sim="{simName}",
th={th},
node_support={nodeSup}
).run()
nm.mnormalize(f="support:_nv",c="range",i=op+"/node.csv").mcal(c="${{_nv}}*200",a="_nodeSize",o=op+"/node2.csv").run()
nnpie.mnetpie(ei=op+"/edge.csv",
ni=op+"/node2.csv",
ef="node1,node2",
nf="node",
o=op+"/graph.html",
nodeSizeFld="_nodeSize",
undirect={undirect}
)
print("#### END")
""".format(**params)
#os.system("fdp)
"""
mgv(ei=op+"/edge.csv",ef="e1,e2",o=op+"/graph.dot")
"""
#edgeWidthFld="{simFld}",
script_w.value = script
return True
def exe(self,script_w):
params={}
params["version"]=self.version
params["date"]=self.date
params["iFile"]= self.iName_w.value
params["oPath"]= self.oPath_w.value
params["oDir"] = self.oDir_w.value
if not wlib.iFileCheck(params["iFile"],self.parent.msg_w):
return False
if not wlib.blankCheck(params["oDir"],"出力dir名",self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"],self.parent.msg_w):
return False
params["tid"] =self.traID_w.getValue()
params["item"] =self.item_w.getValue()
params["klass"] =self.class_w.getValue()
params["pType"] =self.type_w.value[0] # 0は先頭文字の(F,C,M)
params["minSup"]=self.minSup_w.value
params["maxSup"]=self.maxSup_w.value
params["minLen"]=self.minLen_w.value
params["maxLen"]=self.maxLen_w.value
params["minGR"] =self.minGR_w.value
params["top"] =self.top_w.value
if params["item"]==params["tid"]:
self.parent.msg_w.value="##ERROR: トランザクションIDとアイテム項目が同じ項目になってます"
return False
if self.usetop_w.value==True:
script="""
#################################
# mitemset_w.pyの自動生成スクリプト
# version: {version}
# 実行日時: {date}
#################################
import os
import nysol.take as nt
import nysol.mcmd as nm
nm.setMsgFlg(True)
print("#### START")
########### parameter設定
traFile="{iFile}" # トランザクションファイル名
oPath="{oPath}/{oDir}" # 出力ディレクトリ名
tid="{tid}" # トランザクションID項目
item="{item}" # アイテム項目
klass="{klass}" # クラス項目
pType="{pType}" # パターンタイプ(F:頻出集合,C:飽和集合,M:極大集合)
minSup={minSup} # minimum support
maxSup={maxSup} # maximum support
minLen={minLen} # minimum length of items
maxLen={maxLen} # maximum length of items
minGR={minGR} # minimum post-probability (klass指定時のみ有効)
top={top} # 抽出上位件数(supportの大きい順)
os.makedirs(oPath,exist_ok=True)
nt.mitemset(i=traFile,
cls=klass,
tid=tid,
item=item,
O=oPath,
type=pType,
s=minSup,
sx=maxSup,
l=minLen,
u=maxLen,
p=minGR,
top=top).run()
print("#### END")
""".format(**params)
else:
script="""
#################################
# mitemset_w.pyの自動生成スクリプト
# version: {version}
# 実行日時: {date}
#################################
import os
import nysol.take as nt
import nysol.mcmd as nm
nm.setMsgFlg(True)
print("#### START")
########### parameter設定
traFile="{iFile}" # トランザクションファイル名
oPath="{oPath}/{oDir}" # 出力ディレクトリ名
tid="{tid}" # トランザクションID項目
item="{item}" # アイテム項目
klass="{klass}" # クラス項目
pType="{pType}" # パターンタイプ(F:頻出集合,C:飽和集合,M:極大集合)
minSup={minSup} # minimum support
maxSup={maxSup} # maximum support
minLen={minLen} # minimum length of items
maxLen={maxLen} # maximum length of items
minGR={minGR} # minimum post-probability (klass指定時のみ有効)
os.makedirs(oPath,exist_ok=True)
nt.mitemset(i=traFile,
cls=klass,
tid=tid,
item=item,
O=oPath,
type=pType,
s=minSup,
sx=maxSup,
l=minLen,
u=maxLen,
p=minGR).run()
print("#### END")
""".format(**params)
# script tabにセット
script_w.value = script
return True
def exe(self,script_w):
params={}
params["iFile"]= self.iName_w.value
params["oPath"]= self.oPath_w.value
params["oFile"]= self.oFile_w.value
if not wlib.iFileCheck(params["iFile"],self.parent.msg_w):
return False
if not wlib.blankCheck(params["oFile"],"出力ファイル名",self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"],self.parent.msg_w):
return False
key=self.key_w.getValue()
field=self.field_w.getValue()
vFr=self.vFr_w.value
vFrEq=self.vFrEq_w.value
vTo=self.vTo_w.value
vToEq=self.vToEq_w.value
reverse=self.reverse_w.value
rFr="("
rTo=")"
if vFrEq:
rFr="["
if vToEq:
rTo="]"
range_="%s%s,%s%s"%(rFr,vFr,vTo,rTo)
par=[]
if key!="":
par.append("k=\""+key+"\"")
par.append("f=\""+field+"\"")
par.append("c=\""+range_+"\"")
if reverse:
par.append("r=True")
par.append("i=\""+params["iFile"]+"\"")
par.append("o=\""+params["oPath"]+"/"+params["oFile"]+"\"")
header="""
#################################
# mselnum_w.pyの自動生成スクリプト
# version: %s
# 実行日時: %s
#################################
"""%(self.version,self.date)
lib="""
import nysol.mcmd as nm
nm.setMsgFlg(True)
print("#### START")
"""
script="""
f=None
f<<=nm.mselnum(%s)
f.run(msg='on')
print("#### END")
"""%(",".join(par))
script_w.value = header+lib+script
return True
def exe(self, script_w):
params = {}
params["iFile"] = self.iName_w.value
params["oPath"] = self.oPath_w.value
params["oFile"] = self.oFile_w.value
params["key"] = self.key_w.getValue()
params["fld"] = self.field_w.getValue()
params["val"] = self.value_w.value
params["newName"] = self.newfld_w.value
params["reverse"] = self.reverse_w.value
params["reverse_r"] = not params["reverse"]
if not wlib.blankCheck(params["val"], "「値」", self.parent.msg_w):
return False
if not wlib.blankCheck(params["newName"], "「目的変数名」",
self.parent.msg_w):
return False
_head = self.head_w.value
_tail = self.tail_w.value
if _head and _tail:
sub = False
head = False
tail = False
elif _head:
sub = False
head = True
tail = False
elif _tail:
sub = False
head = False
tail = True
else:
sub = True
head = False
tail = False
params["sub"] = sub
params["head"] = head
params["tail"] = tail
script = """
klass=None
klass<<=nm.mselstr(k='{key}',f='{fld}',v='{val}',r={reverse},head={head},tail={tail},sub={sub},i='{iFile}')
klass<<=nm.msetstr(v='1',a='{newName}')
klass<<=nm.mcut(f='{key},{newName}')
klass<<=nm.muniq(k='{key},{newName}')
f=None
f<<=nm.mjoin(k='{key}',m=klass,f='{newName}',n=True,i='{iFile}')
f<<=nm.mselstr(f='{fld}',v='{val}',r={reverse_r},head={head},tail={tail},sub={sub})
f<<=nm.mnullto(f='{newName}',v='0',o='{oPath}/{oFile}')
f.run()
print("#### END")
""".format(**params)
header = """
#################################
# mselstr_w.pyの自動生成スクリプト
# version: %s
# 実行日時: %s
#################################
""" % (self.version, self.date)
lib = """
import nysol.mcmd as nm
nm.setMsgFlg(True)
print("#### START")
"""
output = """
# ファイル出力された結果
oFile="%s"
""" % (params["oFile"])
script_w.value = header + lib + script
return True
def exe(self,script_w):
params={}
params["iFile"]= self.iName_w.value
params["oPath"]= self.oPath_w.value
params["oDir"] = self.oDir_w.value
if not wlib.iFileCheck(params["iFile"],self.parent.msg_w):
return False
if not wlib.blankCheck(params["oDir"],"出力dir名",self.parent.msg_w):
return False
if not wlib.oPathCheck(params["oPath"],self.parent.msg_w):
return False
#params["_id"] =self.id_w.getValue()
params["x"] =list(self.x_w.getValue())
params["algo"] =self.algo_w.value
params["k"] =self.k_w.value
#if params["_id"] in params["x"]:
# self.parent.msg_w.value="##ERROR: idと入力変数が重複指定されています"
# return False
if not wlib.blankCheck(params["k"],"クラスタ数",self.parent.msg_w):
return False
header="""
#################################
# mitemset_w.pyの自動生成スクリプト
# version: %s
# 実行日時: %s
#################################
"""%(self.version,self.date)
script="""
import os
import csv
import numpy as np
import pandas as pd
from sklearn.cluster import KMeans
from yellowbrick.cluster import InterclusterDistance
print("#### START")
########### parameter設定
iFile="{iFile}" # 入力CSVファイル名
x={x} # 入力変数
oPath="{oPath}" # 出力パス
oDir="{oDir}" # 出力ディレクトリ名
algo="{algo}" # クラスタリングアルゴリズム
k={k} # クラスタ数
print("## データセット作成中...")
# 出力ディレクトリを作成する
os.makedirs(oPath+"/"+oDir,exist_ok=True)
# CSVデータをpandas DataFrameに読み込む
df=pd.read_csv(iFile,dtype=str) # とりあえずは全項目str型として読み込む
ds=df.loc[:,x] # 入力変数のみ選択する
ds=ds.astype(float) # 全変数をfloat型に設定する
print("## クラスタリング{algo} 実行中...")
# kmeansモデルの設定と実行
model=KMeans(n_clusters=k)
pred=model.fit_predict(ds.values)
print("## 結果出力")
cluster=pd.DataFrame(pred) # 予測結果をDataFrameに
cluster.columns=["cluster"] # 項目名を設定
out=df.join(cluster) # 元のDataframeに結合する
visualizer = InterclusterDistance(model)
visualizer.fit(ds.values)
visualizer.show(outpath="{oPath}/{oDir}/interclusterDistance.png")
visualizer.show()
# CSVに出力
out.to_csv("{oPath}/{oDir}/predict.csv",index_label=None,encoding="utf-8")
print("#### END")
""".format(**params)
script_w.value = script
return True
