Eno: Compute effective sample size with classical method

In s2dv: A Set of Common Tools for Seasonal to Decadal Verification

Compute effective sample size with classical method

Description

Compute the number of effective samples along one dimension of an array. This effective number of independent observations can be used in statistical/inference tests.
The calculation is based on eno function from Caio Coelho from rclim.txt.

Usage

Eno(data, time_dim = "sdate", na.action = na.pass, ncores = NULL)

Arguments

data	A numeric array with named dimensions.
time_dim	A function indicating the dimension along which to compute the effective sample size. The default value is 'sdate'.
na.action	A function. It can be na.pass (missing values are allowed) or na.fail (no missing values are allowed). See details in stats::acf(). The default value is na.pass.
ncores	An integer indicating the number of cores to use for parallel computation. The default value is NULL.

Value

An array with the same dimension as parameter 'data' except the time_dim dimension, which is removed after the computation. The array indicates the number of effective sample along time_dim.

Examples

set.seed(1)
data <- array(rnorm(800), dim = c(dataset = 1, member = 2, sdate = 4, 
                                 ftime = 4, lat = 10, lon = 10))
na <- floor(runif(40, min = 1, max = 800))
data[na] <- NA
res <- Eno(data)

s2dv documentation built on Oct. 13, 2024, 9:07 a.m.