recommendation

Documentation

Index

• type Similarities
  • func NewSimilarities(entries int) Similarities
  • func NewTopSimilaritiesFromMatrix(m *matrix.CosineLabeledMatrix, top int) Similarities
  • func (m Similarities) Get(index int) []int
  • func (m Similarities) GetAll(indexes []int) []int
  • func (m Similarities) GetAllVector(vector mat.Vector) []int
  • func (m Similarities) ScoredSimilar(cosineMatrix *mat.Dense, vector mat.Vector) matrix.LabeledVector
  • func (m Similarities) Set(index int, entries []int)

Types

type Similarities

type Similarities []*bit.Set

Similarities maps a item to a list of similar items

func NewSimilarities

func NewSimilarities(entries int) Similarities

NewSimilarities creates a new empty similaritites strucuture

func NewTopSimilaritiesFromMatrix

func NewTopSimilaritiesFromMatrix(m *matrix.CosineLabeledMatrix, top int) Similarities

NewTopSimilaritiesFromMatrix creates a filled similaritites structure with max top items per item

func (Similarities) Get

func (m Similarities) Get(index int) []int

Get a list of entries at index

func (Similarities) GetAll

func (m Similarities) GetAll(indexes []int) []int

GetAll returns a merged list of entries at indexes

func (Similarities) GetAllVector

func (m Similarities) GetAllVector(vector mat.Vector) []int

GetAllVector returns a merged list of entries, for items in where vector is set

func (Similarities) ScoredSimilar

func (m Similarities) ScoredSimilar(cosineMatrix *mat.Dense, vector mat.Vector) matrix.LabeledVector

ScoredSimilar creates a vector of recommended items based on a binary vector of already liked items

func (Similarities) Set

func (m Similarities) Set(index int, entries []int)

Set a list of entries at index