Independence tests and benchmarks

[Up] [Top]

Documentation for package ‘knnIndep’ version 2.0

Help Pages

knnIndep-package A package giving the formulas of an exact distribution of ith nearest neighbours and two associated tests for independence
benchmark.patchwork.copula Benchmark function for a new type of non-functional dependence
calculate.power Calculate power at a given significance level
generate.benchmark.data Generating functional dependencies
generate.patchwork.copula Generate data from a non-functional dependence
generate.paths Generate all nearest neighbours distances for one point in a sample
generate.roc Generate ROC curve data
knnIndep A package giving the formulas of an exact distribution of ith nearest neighbours and two associated tests for independence
kr Central probabilty
novelTest.chisq A novel test of independence
novelTest.extreme A novel test of independence
optimise.copula.mi optimize the parameter 'c' of 'generate.patchwork.copula'
parameters Central probabilty
Pc_givena Probability of observing the ith nearest neighbour at distance greater or equal c given the (i-1)th nearest neighbour at distance a
Pc_givena4nn Probability of observing the ith nearest neighbour at distance greater or equal c given the 4 previous nearest neighbours
power.plot Plot power of benchmarked tests of independence
P_ceq Probability of observing the ith nearest neighbour at the same distance or larger as the (i-1)th nearest neighbour
P_cge_aeq Probability of observing the ith nearest neighbour at a distance greater or equal to c and the (i-1)th nearest neighbour was observed at distance a
P_cge_ale Probability of observing the ith nearest neighbour at a distance greater or equal to c and the (i-1)th nearest neighbour was observed at distance smaller or equal a
P_di Probability distribution of the distance to the ith nearest neighbour
roc.plot Plot a ROC
run.tests Run several tests of independence on a benchmark of different functional relationships