| Type: | Package |
| Title: | The Folding Test of Unimodality |
| Version: | 1.0 |
| Description: | The basic algorithm to perform the folding test of unimodality. Given a dataset X (d dimensional, n samples), the test checks whether the distribution of the data are rather unimodal or rather multimodal. This package stems from the following research publication: Siffer Alban, Pierre-Alain Fouque, Alexandre Termier, and Christine Largouët. "Are your data gathered?" In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery Data Mining, pp. 2210-2218. ACM, 2018. <doi:10.1145/3219819.3219994>. |
| Encoding: | UTF-8 |
| License: | GPL-3 |
| LazyData: | true |
| RoxygenNote: | 6.1.0 |
| Suggests: | testthat,MASS,knitr,rmarkdown |
| Imports: | stats |
| VignetteBuilder: | knitr |
| NeedsCompilation: | no |
| Packaged: | 2018-09-19 07:23:46 UTC; asr |
| Author: | Alban Siffer [aut, cre], Amossys [cph, fnd] |
| Maintainer: | Alban Siffer <alban.siffer@irisa.fr> |
| Repository: | CRAN |
| Date/Publication: | 2018-09-30 13:10:12 UTC |
Computes the folding ratio of the input data
Description
Computes the folding ratio of the input data
Usage
folding.ratio(X)
Arguments
X |
nxd matrix (n observations, d dimensions) |
Value
the folding ratio
Examples
X = matrix(runif(n = 1000, min = 0., max = 1.), ncol = 1)
phi = folding.statistics(X)
Computes the folding statistics of the input data
Description
Computes the folding statistics of the input data
Usage
folding.statistics(X)
Arguments
X |
nxd matrix (n observations, d dimensions) |
Value
the folding statsistics
Examples
library(MASS)
mu = c(0,0)
Sigma = matrix(c(1,0.5,1,0.5), ncol = 2)
X = mvrnorm(n = 5000, mu = mu, Sigma = Sigma)
Phi = folding.statistics(X)
Perform the folding test of unimodality
Description
Perform the folding test of unimodality
Usage
folding.test(X)
Arguments
X |
$nxd$ matrix (n observations, d dimensions) |
Value
1 if unimodal, 0 if multimodal
Examples
library(MASS)
n = 10000
d = 3
mu = c(0,0,0)
Sigma = matrix(c(1,0.5,0.5,0.5,1,0.5,0.5,0.5,1), ncol = d)
X = mvrnorm(n = n, mu = mu, Sigma = Sigma)
m = folding.test(X)
Computes the confidence bound for the significance level p
Description
Computes the confidence bound for the significance level p
Usage
folding.test.bound(n, d, p)
Arguments
n |
sample size |
d |
dimension |
p |
significance level (between 0 and 1, the lower, the more significant) |
Value
the confidence bound q (the bounds are 1-q and 1+q)
Examples
n = 2000 # number of observations
d = 2 # 2 dimensional data
p = 0.05 # we want the bound at the level 0.05 (classical p-value)
q = folding.test.bound(n,d,p)
Computes the p-value of the folding test
Description
Computes the p-value of the folding test
Usage
folding.test.pvalue(Phi, n, d)
Arguments
Phi |
the folding statistics |
n |
sample size |
d |
dimension |
Value
the p-value (the lower, the more significant)
Examples
library(MASS)
n = 5000
d = 2
mu = c(0,0)
Sigma = matrix(c(1,0.5,1,0.5), ncol = d)
X = mvrnorm(n = n, mu = mu, Sigma = Sigma)
Phi = folding.statistics(X)
p = folding.test.pvalue(Phi,n,d)
Computes the pivot s_2 (approximate pivot)
Description
Computes the pivot s_2 (approximate pivot)
Usage
pivot.approx(X)
Arguments
X |
nxd matrix (n observations, d dimensions) |
Value
the approximate pivot
Examples
library(MASS)
mu = c(0,0)
Sigma = matrix(c(1,0.5,1,0.5), ncol = 2)
X = mvrnorm(n = 5000, mu = mu, Sigma = Sigma)
Phi = pivot.approx(X)