dataset

package

v0.0.0-...-0d9dfb7 Latest Latest Go to latest Published: Dec 17, 2018 License: Apache-2.0 Imports: 10 Imported by: 0

Details

Valid go.mod file

The Go module system was introduced in Go 1.11 and is the official dependency management solution for Go.
Redistributable license

Redistributable licenses place minimal restrictions on how software can be used, modified, and redistributed.
Tagged version

Modules with tagged versions give importers more predictable builds.
Stable version

When a project reaches major version v1 it is considered stable.
Learn more about best practices

Repository

github.com/milosgajdos83/gosom

Documentation ¶

Index ¶

Constants
func LoadCLS(reader io.Reader) (map[int]int, error)
func LoadCSV(r io.Reader) (*mat.Dense, error)
func LoadLRN(reader io.Reader) (*mat.Dense, error)
func Scale(mx mat.Matrix) *mat.Dense
type DataSet
- func New(dataPath string, clsPath string) (*DataSet, error)
- func (ds *DataSet) Scale() *mat.Dense

Constants ¶

View Source

const (
	// LrnHeaderSize holds the size of LRN file header
	LrnHeaderSize = iota
	// LrnHeaderCols holds number of LRN header columns
	LrnHeaderCols
	// LrnHeaderTypes holds number of LRN types
	LrnHeaderTypes
	// LrnHeaderNames not used
	LrnHeaderNames
	// LrnHeaderRows holds number of LRN header rows
	LrnHeaderRows
)

LRN data format constants

Variables ¶

This section is empty.

Functions ¶

func LoadCLS ¶

func LoadCLS(reader io.Reader) (map[int]int, error)

LoadCLS reads classification information from a .cls file. See the specification here: http://databionic-esom.sourceforge.net/user.html#Classification_files____cls_ The only supported header is the Number of datasets (n)

func LoadCSV ¶

func LoadCSV(r io.Reader) (*mat.Dense, error)

LoadCSV loads data set from the path supplied as a parameter. It returns data matrix that contains particular CSV fields in columns. It returns error if the supplied data set contains corrrupted data or if the data can not be converted to float numbers

func LoadLRN ¶

func LoadLRN(reader io.Reader) (*mat.Dense, error)

LoadLRN reads data from a .lrn file. See the specification here: http://databionic-esom.sourceforge.net/user.html#Data_files____lrn_

func Scale ¶

func Scale(mx mat.Matrix) *mat.Dense

Scale centers the data set to zero mean values in each column and then normalizes them. It does not modify the data stored in the matrix supplied as a parameter.

Types ¶

type DataSet ¶

type DataSet struct {
	Data    *mat.Dense
	Classes map[int]int
}

DataSet represents training data set

func New ¶

func New(dataPath string, clsPath string) (*DataSet, error)

New returns pointer to dataset or fails with error if either the file in dataPath does not exist or if it is encoded in an unsupported format. File format is inferred from the file extension. Currently only csv and lrn data formats are supported. If the dataset has classification information it can be provided as the second parameter. If the file in clsPath doesn't exist New fails with error.

func (*DataSet) Scale ¶

func (ds *DataSet) Scale() *mat.Dense

Scale normalizes data in each column based on its mean and standard deviation and returns it. It modifies the underlying daata. If this is not desirable use the standalone Scale function.

Source Files ¶

View all Source files

dataset.go

?	: This menu
/	: Search site
f or F	: Jump to
y or Y	: Canonical URL