linksAnalysis: Hierarchical clustering to generate final clusters

Description Usage Arguments Value

Description

Given similarities/predictions for each pair, generate transitive closures using hierarchical clustering, returning pairwise match prediction

Usage

1
linksAnalysis(pairs, predCol, myCutoff, myMethod, hash1, hash2)

Arguments

pairs

pairs with 'predCol' column, which are similarity scores/predictions (higher = more similar)

predCol

name of column with similarities/predictions, input as character, e.g. "myPreds"

myCutoff

similarity cutoff (>= myCutoff corresponds to a "match")

myMethod

type of linkage, can be "single", "complete", "average", "minimax"

hash1

name of column with first item in comparison, e.g. "image1"

hash2

name of column with second item in comparison, e.g. "image2"

Value

vector (logical) indicating if pair is matched or not, length is the same as 'pairs'


xhtai/cartridges documentation built on June 1, 2019, 2:58 p.m.