pylearn.classification package

Submodules

pylearn.classification.gaussian_naive_bayes module

class pylearn.classification.gaussian_naive_bayes.GaussianNaiveBayes

Bases: object

Computes continuous classification problems by applying the Bayes theorem with a gaussian distribution.

Attributes:

classes (list):

A list of all classes

mean (numpy.ndarray | pandas.DataFrame):

Mean of all features

variance (numpy.ndarray | pandas.DataFrame):

Variance of all features

prior (numpy.ndarray | pandas.DataFrame):

Prior of Bayes theorem

fit(X: ndarray | DataFrame | Series, Y: ndarray | DataFrame | Series, log_duration=True) → None

Trains the algorithm. Input can be a numpy or pandas object.

Parameters:

X (numpy.ndarray | pandas.DataFrame | pd.Series):

Training input

Y (numpy.ndarray | pandas.DataFrame | pd.Series):

Training output

log_duration (bool, optional):

Logs the duration of the training, default: True

Returns:

None

predict(X: ndarray | DataFrame) → ndarray

Computes the output of a given X.

Parameters:

X (numpy.ndarray | pandas.DataFrame):

Testing input

Returns:

Predicted classes as array

pylearn.classification.multinomial_naive_bayes module

class pylearn.classification.multinomial_naive_bayes.MultinomialNaiveBayes

Bases: object

Computes text classification problems by applying the Bayes theorem.

Attributes:

texts (pandas.Series):

Input texts used for training

classes (list):

List of unique classes in the training data

num_of_samples (int):

Total number of samples in the training data

prior (pandas.DataFrame):

Prior probabilities of each class

vocab (list):

List of unique words in the training data

posterior (pandas.DataFrame):

Posterior probabilities of each word given each class

fit(X: Series, Y: Series, alpha=1, log_duration=True) → None

Trains the algorithm. Data must not be continuous data.

Parameters:

X (pandas.Series):

Training input

Y (pandas.Series):

Training output

alpha (float, optional):

Smoothing parameter

log_duration (bool, optional):

Logs the duration of the training, default: True

Returns:

None

predict(X: Series) → ndarray

Computes the output of a given X.

Parameters:

X (pandas.Series):

Testing input

Returns:

Predicted classes as array

pylearn.classification.temp module

Module contents