generate_smts: Sequential Minimal Training Samples for Categorical Data

In pedro-teles-fonseca/polya: Bayesian Hypothesis Testing for Categorical Data

Description

\loadmathjax

Generate sequential minimal training samples for Binomial and Multinomial models as described in \insertCiteberger2004training;textualpolya.

Usage

generate_smts(x, n = 2 * length(x[!is.na(x)]))

Arguments

x	Atomic vector of an eligible data type (logical, integer, double and character).
n	Numeric vector of length 1 indicating the number of samples to generate.

Details

Training samples play a central role in a variety of statistical methodologies and are particularly usefull to

Sequential minimal training samples (SMTS) are obtained by randomly drawing from x (without replacement for a given SMTS), stopping when the subset is a proper traiing sample.

Value

Returns a list with length n. Each element of this list contains a minimal training sample with data of the same type as x.

References

\insertAllCited

See Also

• ibf, aibf and gibf for the implementation of intrinsic Bayes factors that use sequential minimal training samples.

Examples

set.seed(10)

data_1 <- sample(letters[1:3], size = 5, replace = TRUE)
generate_smts(data_1, 3)

data_2 <- sample(1:10, size = 3, replace = TRUE)
generate_smts(data_2, 5)

pedro-teles-fonseca/polya documentation built on Jan. 30, 2021, 6:47 p.m.