calc_neg_log_lik_vect_multivariate_chunk_outer: The outer function for calculating the negative...

In MichaelHoltonPrice/yada: Yet Another Demographic Analysis package

Description

Use a parallel for loop along with "chunks" of input data to calculate the vector of negative log-likelihoods. For each chunk, a conventional for loop is used for each negative log-likelihood calculation and a parallel for loop is used to iterate across chunks. The output is identical to calc_neg_log_lik_vect_multivariate, but chunking the data can lead to better performance on some syanalysis_names.

Usage

calc_neg_log_lik_vect_multivariate_chunk_outer(
  th_y,
  calc_data,
  tf_cat_vect = NA,
  num_chunks = round(length(calc_data)/10)
)

Arguments

th_y	The parameter vector
calc_data	Data needed for a speedy negative log-likelihood calculation
tf_cat_vect	A vector to transform th_y from an unconstrained to a constrained representation (if necessary)
num_chunks	The number of chunks to use (default: round(length(calc_data)/10))

Value

The vector of negative log-likelihoods

MichaelHoltonPrice/yada documentation built on Sept. 19, 2021, 11:27 p.m.