Two sample Test of Homogeneity

This function allows to perform various nonparametric test for homogeneity on two samples of circular data.

circular_test(x, y, test = "dixon", alpha = 0.05, B = NULL,
  type = "exact", seed = 1982)

Arguments

x	first sample
y	second sample
test	considered test (dixon for Dixon test (default); ww for Wheeler-Watson; wilcox for Wilcoxon test; rao for Rao test; vdw for van der Waerden test; savage for Savage test
alpha	significance level (default = 0.05)
B	number of bootstrap replications
seed	seed used for random number generation
type	method to compute pvalues (available methods: exact for exact computation (default) which is appropriate for small sample sizes; mc for approximation based on Monte-carlo simulations)

Value

A list with the following structure:

cv

a list containing the results of the exact distribution (NULL if type = "mc"), see function get_critical_values for details

mc

a list containing the results of the approximated distribution obtained by simulation (NULL if type = "exact"), see function MC_pvalue for details

B

number of bootstrap replications

alpha

significance level

test

the considered test

stat

observed test statistic

spacings

a list containing the observed spacings, see function compute_Sk for details

x

expression deparsing of the first dataset

y

expression deparsing of the second dataset

Examples

# Load dataset
data(pigeons)
# Dixon test (exact pvalue)
circular_test(pigeons$experimental, pigeons$control)
#> 
#>       Dixon Two Sample Test
#> 
#> Data:  pigeons$experimental and pigeons$control
#> Test Statistic: 41
#> Exact P-value: 0.02096
#> Bracketing Points and Pair of Signif. Levels:
#> c1 = 33 (p1 = 0.0567)
#> c2 = 35 (p2 = 0.0460)

# Dixon test (approximated pvalue)
circular_test( pigeons$experimental, pigeons$control, type = "mc")
#> 
#>       Dixon Two Sample Test
#> 
#> Data:  pigeons$experimental and pigeons$control
#> Test Statistic: 41
#> Approx. P-value: 0.0228
#> P-value stand. error: 0.0009575
#> based on 10000 Monte-Carlo replications