Classes
class	AttributeDescriptor
class	Dataset
Functions
def	bootstrap_samples
	Makes bootstrap samples.
def	dataset_entropy_discrete
	Calculates entropy in a dataset's output labels.
def	entropy_discrete
	Caculates the discrete entropy of a data vector.
def	leave_one_out
	Good for leave one out cross validation loops.
def	split_continuous
	Splits up a dataset based on value in a particular attribute.
def	unique_values
	Common quantities calculated from datasets.

Function Documentation

def ml_lib.dataset.bootstrap_samples	(	dataset,
		number_samples,
		points_per_sample
	)

Makes bootstrap samples.

Parameters:

dataset	Dataset object
number_samples	number of bootstrap set to generate
points_per_sample	number of points in each sample

Definition at line 94 of file dataset.py.

def ml_lib.dataset.dataset_entropy_discrete ( dataset )

Calculates entropy in a dataset's output labels.

Parameters:

dataset

Definition at line 146 of file dataset.py.

def ml_lib.dataset.entropy_discrete ( data )

Caculates the discrete entropy of a data vector.

Parameters:

data	1xn matrix of discrete values

Definition at line 132 of file dataset.py.

def ml_lib.dataset.leave_one_out	(	dataset,
		index
	)

Good for leave one out cross validation loops.

Parameters:

dataset
index

Definition at line 52 of file dataset.py.

def ml_lib.dataset.split_continuous	(	dataset,
		attribute,
		split_point
	)

Splits up a dataset based on value in a particular attribute.

Parameters:

attribute	attribute to split on
split_point	value in that attribute to split on

Definition at line 63 of file dataset.py.

def ml_lib.dataset.unique_values	(	data,
		attribute_number = `0`
	)

Common quantities calculated from datasets.

Returns unique values represented by an attribute

ex. unique_values(np.matrix([1, 2, 3, 4, 4, 4, 5]), 0) returns [1,2,3,4,5]

Parameters:

data	nxm matrix where each column is a data vector
attribute_number	row to find unique values in

Definition at line 120 of file dataset.py.

Classes