def generateViews(self):
T=map(list,zip(*self.D))
if self.transformed:
for column_id in range(self.column_num):
f = Features(self.names[column_id],self.types[column_id],self.origins[column_id])
#calculate min,max for numerical
if f.type==Type.numerical:
if self.classify_num==1 or not self.describe2:#not categorized or categorized scatter
f.min,f.max=min(T[column_id]),max(T[column_id])
f.minmin=f.min
if f.min==f.max:
self.types[column_id]=f.type=Type.none
self.features.append(f)
continue
else:
delta=self.tuple_num/self.classify_num
f.min=[min(T[column_id][class_id*delta:(class_id+1)*delta]) for class_id in range(self.classify_num)]
f.minmin=min(f.min)
f.max=[max(T[column_id][class_id*delta:(class_id+1)*delta]) for class_id in range(self.classify_num)]
if sum([f.max[class_id]-f.min[class_id] for class_id in range(self.classify_num)])==0:
self.types[column_id]=f.type=Type.none
self.features.append(f)
continue
if min(f.min)==max(f.min) and min(f.max)==max(f.max):
if sum([0 if T[column_id][class_id*delta:(class_id+1)*delta]==T[column_id][(class_id+1)*delta:(class_id+2)*delta] else 1 for class_id in range(self.classify_num-1)])==0:
self.types[column_id]=f.type=Type.none
self.features.append(f)
continue
#calculate distinct,ratio for categorical,temporal
if f.type==Type.categorical or f.type==Type.temporal:
f.distinct=self.tuple_num
f.ratio=1.0
self.features.append(f)
else:
for column_id in range(self.column_num):
f = Features(self.names[column_id],self.types[column_id],self.origins[column_id])
#calculate min,max for numerical,temporal
if f.type==Type.numerical or f.type==Type.temporal:
f.min,f.max=min(T[column_id]),max(T[column_id])
f.minmin=f.min
if f.min==f.max:
self.types[column_id]=f.type=Type.none
self.features.append(f)
continue
d={}
#calculate distinct,ratio for categorical,temporal
if f.type == Type.categorical or f.type == Type.temporal:
for i in range(self.tuple_num):
if self.D[i][column_id] in d:
d[self.D[i][column_id]]+=1
else:
d[self.D[i][column_id]]=1
f.distinct = len(d)
if f.distinct==1:
self.types[column_id]=f.type=Type.none
self.features.append(f)
continue
f.ratio = 1.0 * f.distinct / self.tuple_num
f.distinct_values=[(k,d[k]) for k in sorted(d)]
if f.type==Type.temporal:
self.getIntervalBins(f)
self.features.append(f)
#generate 2D views
if self.describe2=='' and self.classify_id==-1:
for i in range(self.column_num):
for j in range(self.column_num):
if i==j:
continue
fi=self.features[i]
fj=self.features[j]
if fi.type==Type.categorical and fj.type==Type.numerical and fi.ratio==1.0:
charts=[]
if fj.minmin>0 and fi.distinct<=5 and not (len(fj.name)>=6 and fj.name[0:4]=='AVG(' and fj.name[-1]==')'):
charts.append(Chart.pie)
if fi.distinct<=20:
charts.append(Chart.bar)
elif fi.type==Type.temporal and fj.type==Type.numerical and fi.ratio==1.0:
charts=[]
if fi.distinct<7:
charts.append(Chart.bar)
else:
charts.append(Chart.line)
elif (not self.transformed) and fi.type==Type.numerical and fj.type==Type.numerical and i<j:
charts=[Chart.scatter]
else:
charts=[]
for chart in charts:
v=View(self,i,j,-1,1,[T[i]],[T[j]],chart)
self.views.append(v)
self.view_num+=1
#generate 3D views
elif self.describe2:
for i in range(self.column_num):
for j in range(self.column_num):
fi=self.features[i]
fj=self.features[j]
if fi.type==Type.categorical and fj.type==Type.numerical and fj.minmin>0:
charts=[Chart.bar]
elif fi.type==Type.temporal and fj.type==Type.numerical:
if self.tuple_num/self.classify_num<7:
charts=[Chart.bar]
else:
charts=[Chart.line]
else:
charts=[]
for chart in charts:
delta=self.tuple_num/self.classify_num
series_data = [T[j][series * delta:(series + 1) * delta] for series in range(self.classify_num)]
v = View(self, i, j, self.classify_id, self.classify_num, [T[i][0:delta]], series_data, chart)
self.views.append(v)
self.view_num += 1
else:
for i in range(self.column_num):
for j in range(self.column_num):
if i>=j or self.types[i]!=Type.numerical or self.types[j]!=Type.numerical:
continue
X=[]
Y=[]
id=0
for k in range(self.classify_num):
x=T[i][id:id+self.classes[k][1]]
y=T[j][id:id+self.classes[k][1]]
id+=self.classes[k][1]
X.append(x)
Y.append(y)
v=View(self,i,j,self.classify_id,self.classify_num,X,Y,Chart.scatter)
self.views.append(v)
self.view_num+=1
self.instance.view_num+=self.view_num
def getFeatures(self):
self.T = map(list, zip(*self.D))
if self.transformed == True:
for column_id in range(self.column_num):
f = Features(self.names[column_id], self.types[column_id],
self.origins[column_id])
# calculate min,max for numerical
if f.type == Type.numerical:
if self.classify_num == 1 or not self.describe2: # not categorized or categorized scatter
f.min, f.max = min(self.T[column_id]), max(
self.T[column_id])
f.minmin = f.min
if f.min == f.max:
self.types[column_id] = f.type = Type.none
self.features.append(f)
continue
else:
delta = self.tuple_num / self.classify_num
f.min = [
min(self.T[column_id][class_id *
delta:(class_id + 1) *
delta])
for class_id in range(self.classify_num)
]
f.minmin = min(f.min)
f.max = [
max(self.T[column_id][class_id *
delta:(class_id + 1) *
delta])
for class_id in range(self.classify_num)
]
if sum([
f.max[class_id] - f.min[class_id]
for class_id in range(self.classify_num)
]) == 0:
self.types[column_id] = f.type = Type.none
self.features.append(f)
continue
if min(f.min) == max(f.min) and min(f.max) == max(
f.max):
if sum([
0
if self.T[column_id][class_id *
delta:(class_id + 1) *
delta]
== self.T[column_id][(class_id + 1) *
delta:(class_id + 2) *
delta] else 1
for class_id in range(self.classify_num -
1)
]) == 0:
self.types[column_id] = f.type = Type.none
self.features.append(f)
continue
# calculate distinct,ratio for categorical,temporal
if f.type == Type.categorical or f.type == Type.temporal:
f.distinct = self.tuple_num
f.ratio = 1.0
self.features.append(f)
else:
for column_id in range(self.column_num):
f = Features(self.names[column_id], self.types[column_id],
self.origins[column_id])
#calculate min,max for numerical,temporal
if f.type == Type.numerical or f.type == Type.temporal:
f.min, f.max = min(self.T[column_id]), max(
self.T[column_id])
f.minmin = f.min
if f.min == f.max:
self.types[column_id] = f.type = Type.none
self.features.append(f)
continue
d = {}
#calculate distinct,ratio for categorical,temporal
if f.type == Type.categorical or f.type == Type.temporal:
for i in range(self.tuple_num):
if self.D[i][column_id] in d:
d[self.D[i][column_id]] += 1
else:
d[self.D[i][column_id]] = 1
f.distinct = len(d)
f.ratio = 1.0 * f.distinct / self.tuple_num
f.distinct_values = [(k, d[k]) for k in sorted(d)]
if f.type == Type.temporal:
self.getIntervalBins(f)
self.features.append(f)
def which_chart(tables,table_num):
i=1
while i< table_num:
charts = []
self=tables[i] # type: Table
T=map(list,zip(*self.D))
if self.transformed:
for column_id in range(self.column_num):
f = Features(self.names[column_id],self.types[column_id],self.origins[column_id])
#calculate min,max for numerical
if f.type==Type.numerical:
if self.classify_num==1 or not self.describe2:#not categorized or categorized scatter
f.min,f.max=min(T[column_id]),max(T[column_id])
f.minmin=f.min
if f.min==f.max:
self.types[column_id]=f.type=Type.none
self.features.append(f)
continue
else:
delta=self.tuple_num/self.classify_num
f.min=[min(T[column_id][class_id*delta:(class_id+1)*delta]) for class_id in range(self.classify_num)]
f.minmin=min(f.min)
f.max=[max(T[column_id][class_id*delta:(class_id+1)*delta]) for class_id in range(self.classify_num)]
if sum([f.max[class_id]-f.min[class_id] for class_id in range(self.classify_num)])==0:
self.types[column_id]=f.type=Type.none
self.features.append(f)
continue
if min(f.min)==max(f.min) and min(f.max)==max(f.max):
if sum([0 if T[column_id][class_id*delta:(class_id+1)*delta]==T[column_id][(class_id+1)*delta:(class_id+2)*delta] else 1 for class_id in range(self.classify_num-1)])==0:
self.types[column_id]=f.type=Type.none
self.features.append(f)
continue
#calculate distinct,ratio for categorical,temporal
if f.type==Type.categorical or f.type==Type.temporal:
f.distinct=self.tuple_num
f.ratio=1.0
self.features.append(f)
else:
print "error:table"+str(i)+"not transformed"
for j in range(self.column_num):
for k in range(self.column_num):
fj=self.features[j]
fk=self.features[k]
if fj.name[0:4]=='CNT(' and fj.name[-1]==')'and fk.ratio==1.0:
if fk.type==Type.numerical:
charts = []
charts.append(Chart.scatter)
charts.append(Chart.bar)
charts.append(Chart.pie)
if fk.type==Type.categorical:
charts = []
charts.append(Chart.bar)
charts.append(Chart.pie)
elif fk.type==Type.temporal:
charts = []
charts.append(Chart.bar)
charts.append(Chart.line)
# charts.append(Chart.pie)
else:
charts = []
for chart in charts:
v=View(self,k,j,-1,1,[T[k]],[T[j]],chart)
self.countviews.append(v)
self.countview_num+=1
if charts:
while self.countview_num!=1:
if self.countviews[1]:
if self.countviews[0].M<self.countviews[1].M:
del self.countviews[0]
self.countview_num-=1
else:
del self.countviews[1]
self.countview_num-=1
i+=1
return tables
def generateViews(self):
T=map(list,zip(*self.D)) # the '*' is for unzipping self.D
if self.transformed:
for column_id in range(self.column_num):
f = Features(self.names[column_id],self.types[column_id],self.origins[column_id])
#calculate min,max for numerical
if f.type==Type.numerical:
f.min,f.max=min(T[column_id]),max(T[column_id])
if f.min==f.max:
self.types[column_id]=f.type=Type.none
self.features.append(f)
continue # directly go to the next value in for loop
#calculate distinct,ratio for categorical,temporal
if f.type==Type.categorical or f.type==Type.temporal:
f.distinct=self.tuple_num
f.ratio=1.0
self.features.append(f)
else:
for column_id in range(self.column_num):
f = Features(self.names[column_id],self.types[column_id],self.origins[column_id])
#calculate min,max for numerical,temporal
if f.type==Type.numerical or f.type==Type.temporal:
f.min,f.max=min(T[column_id]),max(T[column_id])
if f.min==f.max:
self.types[column_id]=f.type=Type.none
self.features.append(f)
continue
d={}
#calculate distinct,ratio for categorical,temporal
if f.type == Type.categorical or f.type == Type.temporal:
for i in range(self.tuple_num):
if self.D[i][column_id] in d:
d[self.D[i][column_id]]+=1
else:
d[self.D[i][column_id]]=1
f.distinct = len(d)
f.ratio = 1.0 * f.distinct / self.tuple_num
f.distinct_values=[(k,d[k]) for k in sorted(d)]
if f.type==Type.temporal:
self.getIntervalBins(f)
self.features.append(f)
#generate 2D views
if self.describe2=='' and self.classify_id==-1:
for i in range(self.column_num):
for j in range(self.column_num):
if i==j: # all combinations of 2 columns except the same column
continue
fi=self.features[i]
fj=self.features[j]
if fi.type==Type.categorical and fj.type==Type.numerical and fi.ratio==1.0:
charts=[]
if fj.min>0 and fi.distinct<=5 and not (len(fj.name)>=6 and fj.name[0:4]=='AVG(' and fj.name[-1]==')'): # AVG makes no sense in pie chart
charts.append(Chart.pie)
if fi.distinct<=20:
charts.append(Chart.bar)
elif fi.type==Type.temporal and fj.type==Type.numerical and fi.ratio==1.0:
charts = []
'''if fj.min>0 and fi.distinct<=5 and not (len(fj.name)>=6 and fj.name[0:4]=='avg(' and fj.name[-1]==')'):
charts.append(Chart.pie)'''
if fi.distinct < 7:
charts.append(Chart.bar)
else:
charts.append(Chart.line)
elif (not self.transformed) and fi.type==Type.numerical and fj.type==Type.numerical and i<j:
charts=[Chart.scatter]
else:
charts=[]
for chart in charts:
v=View(self,i,j,-1,1,[T[i]],[T[j]],chart) # the function to visualize the prepared table
self.views.append(v)
self.view_num+=1
#generate 3D views
elif self.describe2:
for i in range(self.column_num):
for j in range(self.column_num):
fi=self.features[i]
fj=self.features[j]
if fi.type==Type.categorical and fj.type==Type.numerical and fj.min>0:
charts=[Chart.bar]
elif fi.type==Type.temporal and fj.type==Type.numerical:
if self.tuple_num/self.classify_num<7:
charts=[Chart.bar]
else:
charts=[Chart.line]
else:
charts=[]
for chart in charts:
delta=self.tuple_num/self.classify_num
series_data = [T[j][series * delta:(series + 1) * delta] for series in range(self.classify_num)]
v = View(self, i, j, self.classify_id, self.classify_num, [T[i][0:delta]], series_data, chart)
self.views.append(v)
self.view_num += 1
else:
for i in range(self.column_num):
for j in range(self.column_num):
if i>=j or self.types[i]!=Type.numerical or self.types[j]!=Type.numerical:
continue
X=[]
Y=[]
id=0
for k in range(self.classify_num):
x=T[i][id:id+self.classes[k][1]]
y=T[j][id:id+self.classes[k][1]]
id+=self.classes[k][1]
X.append(x)
Y.append(y)
v=View(self,i,j,self.classify_id,self.classify_num,X,Y,Chart.scatter)
self.views.append(v)
self.view_num+=1
self.instance.view_num+=self.view_num
