HasGP.Data.Normalise

Description

Normalise is a module in the HasGP Gaussian process library. It contains functions for performing basic normalisation tasks on training examples, and for computing assorted standard statistics.

Synopsis

Documentation

exampleMean Source

Arguments

:: Inputs	Matrix - one row per example
-> DVector	Vector of means for each attribute.

Compute the mean for each attribute in a set of examples.

exampleVariance Source

Arguments

:: Inputs	Matrix - one row per example
-> DVector	Vector of variances for each attribute.

Compute the variance for each attribute in a set of examples.

exampleMeanVariance Source

Arguments

:: Inputs	Matrix - one row per example
-> (DVector, DVector)	Means and variances

Compute the mean and variance for each attribute in a set of examples.

normaliseMeanVariance Source

Arguments

:: DVector	Vector of new means required
-> DVector	Vector of new variances required
-> Inputs	Matrix - one row per example
-> Inputs	Normalised matrix

Normalise a set of examples to have specified mean and variance.

normaliseMeanVarianceSimple Source

Arguments

:: Double	New mean required
-> Double	New variance required
-> Inputs	Matrix - one row per example
-> Inputs	Normalised matrix

The same as normaliseMeanVariance but every column (attribute) is normalised in the same way.

normaliseBetweenLimits Source

Arguments

:: Double	New min required
-> Double	New max required
-> Inputs	Matrix - one row per example
-> Inputs	Normalised matrix

Normalise a set of examples to have specified maximum and minimum.

findRedundantAttributes Source

Arguments

:: Inputs	Matrix - one row per example
-> [Bool]	List - True elements mark redundancy

Find the columns of a matrix in which all values are equal.

listRedundantAttributes Source

Arguments

:: Inputs	Matrix - one row per example
-> [Int]	List - positions of redundant attributes

List column numbers for redundant attributes.

removeRedundantAttributes Source

Arguments

:: Inputs	Matrix - one row per example
-> Inputs	Modified matrix - one row per example

Remove any redundant columns from a matrix.

retainAttributes Source

Arguments

:: [Int]	List of columns to keep.
-> Inputs	Matrix - one row per example
-> Inputs	Modified matrix - one row per example

Specify a list of columns (matrix numbered from 1). Produce a matrix with ONLY those columns in the order specified in the list.

confusionMatrix :: Targets -> Outputs -> (Double, Double, Double, Double)Source

Compute the numbers for the confusion matrix. It is assumed that classes are +1 (positive) and -1 (negative). Result is (a,b,c,d): a - correct negatives b - predict positive when correct is negative c - predict negative when correct is positive d - correct positives

printConfusionMatrix Source

Arguments

:: Targets	Vector of targets
-> Outputs	Vector of actual outputs
-> IO ()

Print the confusion matrix and some other statistics

countLabels :: Targets -> IO ()Source

Assuming the labels are +1 or -1, count how many there are of each.