from utilities import ConfusionMatrix
a = [0, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3]
p = [0, 1, 1, 1, 1, 2, 2, 2, 2, 2, 3, 1, 1, 3, 3, 3]
cm = ConfusionMatrix(a, p)
print("ACTUAL")
print(a)
print("PREDICTED")
print(p)
for i in range(4):
print("\ncategory %s:" % i)
print("TP: ", end="")
print(cm.TP(label=i))
print("FP: ", end="")
print(cm.FP(label=i))
print("TN: ", end="")
print(cm.TN(label=i))
print("FN: ", end="")
print(cm.FN(label=i))
print("avg accuracy: ", end="")
print(cm.average_accuracy())
print("accuracy: ", end="")
print(cm.accuracy())
print("precision: ", end="")
print(cm.precision())
print("recall: ", end="")
print(cm.recall())
return Y_test, predicted
to_save = []
if test_what == 'train_size':
to_save.append(('train_size', 'accuracy', 'precision', 'recall'))
for train_size in [ 1000*n for n in range(1,10) ]:
print('train_size:',train_size)
# train the naive bayes and obtain the actual, predicted vectors.
actual, predicted = run_nb(train_size=train_size, learn_code=learn_code)
# get confusion matrix to get metrics
CM = ConfusionMatrix(actual, predicted)
to_save.append( (train_size, CM.average_accuracy(), CM.precision(), CM.recall() ) )
elif test_what == 'cumulative_ngram':
to_save.append(('ngram_max', 'accuracy', 'precision', 'recall'))
for max_ngram in range(4,9):
print('max_ngram:',max_ngram)
# train the naive bayes and obtain the actual, predicted vectors.
actual, predicted = run_nb(ngram_range=(1,max_ngram), learn_code=learn_code)
# get confusion matrix to get metrics
CM = ConfusionMatrix(actual, predicted)
to_save.append( (max_ngram, CM.average_accuracy(), CM.precision(), CM.recall() ) )
to_save = []
if test_what == 'train_size':
to_save.append(('train_size', 'accuracy', 'precision', 'recall'))
for train_size in [1000 * n for n in range(1, 10)]:
print('train_size:', train_size)
# train the naive bayes and obtain the actual, predicted vectors.
actual, predicted = run_nb(train_size=train_size,
learn_code=learn_code)
# get confusion matrix to get metrics
CM = ConfusionMatrix(actual, predicted)
to_save.append(
(train_size, CM.average_accuracy(), CM.precision(), CM.recall()))
elif test_what == 'cumulative_ngram':
to_save.append(('ngram_max', 'accuracy', 'precision', 'recall'))
for max_ngram in range(4, 9):
print('max_ngram:', max_ngram)
# train the naive bayes and obtain the actual, predicted vectors.
actual, predicted = run_nb(ngram_range=(1, max_ngram),
learn_code=learn_code)
# get confusion matrix to get metrics
CM = ConfusionMatrix(actual, predicted)
# In[166]:
mb.fit(X_train2, list(Y_train))
# In[167]:
X_test2 = kb.transform(cv.transform(X_test))
X_test2 = cv.transform(X_test)
# In[168]:
Y_predicted = mb.predict(X_test2)
# In[169]:
cm = ConfusionMatrix(Y_test, Y_predicted)
# In[170]:
cm.average_accuracy()
# In[171]:
cm.confusion_matrix
# In[152]:
df = pd.read_csv('./clean/ml_dataset_test_in-1111.csv', index_col=0)
# In[153]:
# In[167]:
X_test2 = kb.transform(cv.transform(X_test))
X_test2 = cv.transform(X_test)
# In[168]:
Y_predicted = mb.predict(X_test2)
# In[169]:
cm = ConfusionMatrix(Y_test, Y_predicted)
# In[170]:
cm.average_accuracy()
# In[171]:
cm.confusion_matrix
# In[152]:
df = pd.read_csv('./clean/ml_dataset_test_in-1111.csv', index_col=0)
