from sklearn.externals import joblib
from sklearn.metrics import confusion_matrix
from zzz.chapter06.knock52 import load_data
def generate_confusion_matrix(model, x, y):
y_pred = model.predict(x)
c_matrix = confusion_matrix(y, y_pred)
return c_matrix
if __name__ == '__main__':
clf = joblib.load('linear_model.pkl')
(x_train, y_train) = load_data('train{}.txt')
(x_val, y_val) = load_data('valid{}.txt')
train_matrxi = generate_confusion_matrix(clf, x_train, y_train)
print(train_matrxi)
val_matrix = generate_confusion_matrix(clf, x_val, y_val)
print(val_matrix)
from sklearn.externals import joblib
from zzz.chapter06.knock52 import load_data
if __name__ == '__main__':
clf = joblib.load('linear_model.pkl')
(x_val, y_val) = load_data('valid{}.txt')
x = x_val.sample(n=10)
# [b, e, m, t]
res = clf.predict(x)
prob = clf.predict_proba(x)
print('Category: X', '|\tprobability:', ['b', 'e', 'm', 't'])
for (r, p) in zip(res, prob):
print('Category:', r, '|\tprobability:', p)