Python分类器Sklearn

Question

I am quite new to Python and SKLearn. I am trying to make a simple classifier but I am running into a problem. I have been following a few different tutorials but getting an error when I try to use the .fit method. I am new to the concept and have tried the documentation but have found that hard to understand, can any one help with me error or point me in the right direction.

My thinking behind the error is that the values are out of range for the dtype, as I have transformed all the missing values or nan values but the error is still arising

Code

def main():
setup_files()

imputer = Imputer()

#the training data minus id and type:
t_num_data = load_csv(training_set_file_path, range(1, 17))
t_num_data_imputed = imputer.fit_transform(t_num_data)
print(t_num_data_imputed)

#the training type column
t_type_col = load_csv(training_set_file_path, 17, dtype=np.dtype((str, 5)))
#the query data minus id and type:
q_data = load_csv(queries_file_path, range(1, 17))
#the query id column
q_id = load_csv(queries_file_path, 0, dtype=np.dtype((str, 10)))


#fit data above to DTC and predict import
model = tree.DecisionTreeClassifier(criterion='entropy')
model.fit_transform(t_num_data, t_type_col)
predictions = model.predict(q_data)


#output the predictions:
with open(solutions_file_path, 'w') as f:
    for i in range(len(predictions)):
        f.write("{},{}\n".format(q_id[i], predictions[i]))


#fit data above to DTC and predict import
model = tree.DecisionTreeClassifier(criterion='entropy')
model.fit(t_num_data, t_type_col)
predictions = model.predict(q_data)


#output the predictions:
with open(solutions_file_path, 'w') as f:
    for i in range(len(predictions)):
        f.write("{},{}\n".format(q_id[i], predictions[i]))

Error

Traceback (most recent call last):
  File "/Users/Rory/Desktop/classifier.py", line 71, in <module>
main()
  File "/Users/Rory/Desktop/classifier.py", line 60, in main
model.fit_transform(t_num_data, t_type_col)
  File "/Users/Rory/anaconda/lib/python2.7/site-packages/sklearn/base.py", line 458, in fit_transform
return self.fit(X, y, **fit_params).transform(X)
  File "/Users/Rory/anaconda/lib/python2.7/site-packages/sklearn/tree/tree.py", line 154, in fit
    X = check_array(X, dtype=DTYPE, accept_sparse="csc")
  File "/Users/Rory/anaconda/lib/python2.7/site-packages/sklearn/utils/validation.py", line 398, in check_array
_assert_all_finite(array)
  File "/Users/Rory/anaconda/lib/python2.7/site-packages/sklearn/utils/validation.py", line 54, in _assert_all_finite
" or a value too large for %r." % X.dtype)
ValueError: Input contains NaN, infinity or a value too large for dtype('float32').

Answer 1

The problem is your NaN values. There is a long list of ways to estimate NaNs. You could try:

t_num_data.fillna(0)

Which will fill all the missing values with 0, and then your classifier will work, but may not be very accurate. There additional methods that take the mean, estimation based on nearest neighbors, etc. But that should get your code working for now.