def selectRowsForCmat( P , c , seed=None): # Plen=P.__len__() C_sel=set() selected = 0 while C_sel.__len__() < c : C_sel.add(ml.getRandomFromDist(P)) return(C_sel)
def computePvectForRows(fname, beta=1):#, header=True, rowLabels=True) : columns,rows = ml.data_size(fname) ARowMags = [0]*rows P = [0]*rows with open(fname,'rb') as f : reader=csv.reader(f,delimiter=' ') for j,row in enumerate(reader) : if j!=0 : # print(j) j -= 1 for i,value in enumerate(row) : if i!=0: ARowMags[j] += float(value)**2 # print(ARowMags) ARowMagsSum= sum(ARowMags) for k in range(rows): P[k]= beta*ARowMags[k]/( ARowMagsSum ) return(P)
def buildCfromA(fname, c, beta=1 ) :
cols,rows= ml.data_size(fname)
P=computePvectForRows(fname, beta=1)#, header=True, rowLabels=True)
print('Pvector done computing')
C_sel = selectRowsForCmat(P, c)
print('c rows selected')
C=np.zeros([c,cols])
with open(fname,'rb') as f :
reader=csv.reader(f,delimiter=' ')
cnt=0
for j,row in enumerate(reader) :
# print(row)
if j==0 :
pass
elif j-1 in C_sel :
# print(C[cnt,:])
# print(row[1:])
C[cnt,:]=(row[1:])
cnt+=1
return(C)
