import numpy as np
import svm
# Size of vocabulary
voca = 0
with open("../data/vocabularyNN.txt",'r') as f:
for line in f:
voca += 1
# Load training data and label
data = np.loadtxt("../data/dataNN.txt", delimiter=' ', dtype=int)
label = np.loadtxt("../data/labelNN.txt", delimiter=' ', dtype=int)
folds = 11
predictAccCV = []
for k in range(2,folds+1):
predictAcc = svm.svmCV(voca, data, label, k)
predictAccCV.append(predictAcc)
print predictAcc
accuracy = np.array(predictAccCV)
np.save('accNNsvm.npy', accuracy)
import numpy as np
import svm
# Size of vocabulary
voca = 0
with open("../data/vocabulary.txt", 'r') as f:
for line in f:
voca += 1
# Load training data and label
data = np.loadtxt("../data/data.txt", delimiter=' ', dtype=int)
label = np.loadtxt("../data/label.txt", delimiter=' ', dtype=int)
folds = 2
predictAccCV = []
for k in range(2, folds + 1):
predictAcc = svm.svmCV(voca, data, label, k)
predictAccCV.append(predictAcc)
print predictAcc
accuracy = np.array(predictAccCV)
np.save('accAllsvm.npy', accuracy)
import numpy as np
import svm
filetype = 'Adj'
# filetype = 'NN'
# filetype = 'AllWords'
groups = 5
folds = 2
C = [1,10,100,1000]
kernel = ['linear','poly','rbf']
predictAccCV = np.zeros([4,3,folds])
for i,cp in enumerate(C):
for j,ke in enumerate(kernel):
print 'C:',cp
print 'kernel:',ke
predictAccCV[i,j] = svm.svmCV(folds, filetype, groups, cp, ke)
print 'predictAccCV:',predictAccCV
np.save('accSVM'+filetype+'.npy', predictAccCV)
