subrank (version 0.9.9.3)

corc0: Function to estimate copula using ranks and sub-sampling, minimal version.

Description

Minimal version of function corc.

Usage

corc0(datavector,sampsize,dimension,subsampsize,nboot,u,mixties=FALSE,nthreads=2)

Value

the number of hits for each vector of ranks, plus 2 last values of the vector : number of ties and number of sub-samples really used.

Arguments

datavector

a vector, containing the observations

sampsize

the sample size

dimension

the sample dimension

subsampsize

the sub-sample size

nboot

the number of sub-samples (must be big)

u

a random seed, integer

mixties

if TRUE, put equal weight on tied values, using random permutations

nthreads

number of number of threads, assumed to be strictly positive. For "full throttle" computations, consider using parallel::detectCores()

Author

Jerome Collet

Examples

lon <- 30
a <- 2.85
x <- rnorm(lon)
y = a*x^2+rnorm(lon)
c=corc0(c(x,y),lon,2,8,1e5,75014)
c

c0=c(
1203, 1671, 1766, 959, 1586, 1715, 1803, 1205, 1260,1988, 2348, 1917, 3506, 2045, 1340,
1093, 2694, 2757,2233, 1085, 2322, 1793, 1569, 1263, 1709, 1747, 1512,1308, 1778, 1354,
1184, 1097, 2487, 2730, 2112, 1100,2435, 2033, 1572, 1093, 1369, 1722, 1462, 1015, 1228,
1419, 1776, 1852, 1009, 1097, 1179, 1323, 1595, 1316,1477, 2628, 889, 1178, 1981, 4000, 
35, 840, 2091, 4467,0, 27405)
set.seed(75013)
lon=30
dimension=3
sssize=4
c0==corc0(rnorm(lon*dimension),lon,dimension,sssize,1e5,75014)