def testSHM():
threshold = 150
#Load data
pkl_file = open('data.plk', 'rb')
pos, neg, Ni, Nr = pickle.load(pkl_file)
idx = np.random.permutation(len(Ni))
train_len = round(len(Ni) * 0.8)
train_idx = idx[:train_len]
test_idx = idx[train_len:]
x_train = [Ni[i] for i in train_idx]
y_train = [Nr[i] for i in train_idx]
x_test = [Ni[i] for i in test_idx]
y_test = [Nr[i] for i in test_idx]
#CS model test
cs = CS()
cs.train(x_train, y_train)
cs.predict(x_test)
print "RSE: %f" % cs.RSE(y_test)
print "precision: %f, recall: %f" % cs.PR(y_test,threshold)
#LN model test
ln = LN()
ln.train(x_train, y_train)
ln.predict(x_test)
print "RSE: %f" % ln.RSE(y_test)
print "precision: %f, recall: %f" % ln.PR(y_test,threshold)
#CreatData('../../corpus_g2','data.plk',threshold) #just run once the pos: 425, the neg: 682
#Load data
pkl_file = open('sequence.plk', 'rb')
_ ,time = pickle.load(pkl_file)
idx = np.random.permutation(len(Ni))
train_len = round(len(Ni) * 0.8)
train_idx = idx[:train_len]
test_idx = idx[train_len:]
x_train = [Ni[i] for i in train_idx]
y_train = [Nr[i] for i in train_idx]
x_test = [Ni[i] for i in test_idx]
y_test = [Nr[i] for i in test_idx]
#CS model test
cs = CS()
cs.train(x_train, y_train)
cs.predict(x_test)
print "RSE: %f" % cs.RSE(y_test)
print "precision: %f, recall: %f" % cs.PR(y_test,threshold)
#LN model test
ln = LN()
ln.train(x_train, y_train)
ln.predict(x_test)
print "RSE: %f" % ln.RSE(y_test)
print "precision: %f, recall: %f" % ln.PR(y_test,threshold)
print 'exit'
