log_likelihood: Calculate log likelihood (LLK)

In W-Mohammed/calibR: Calibration of Decision-Analytic Models

Calculate log likelihood (LLK)

Description

Calculate log likelihood (LLK)

Usage

log_likelihood(.samples, .func, .args, .l_targets)

Arguments

.samples	A table or vector of sampled parameter values
.func	A function defining the model to be calibrated
.args	A list of arguments to be passed to .func
.l_targets	A list containing a vector of targets' names, a vector of targets' weights, a vector of targets' distributions, and a table for each target that contains the values (column name 'value') and standard errors (column name 'sd') of the corresponding target.

Value

A table with proposed parameter sets and their corresponding summed overall likelihood values sorted in descending order.

Examples

## Not run: 
library(calibR)
data("CRS_targets")
Surv <- CRS_targets$Surv
v_targets_names <- c("Surv")
v_targets_dists <- c('norm')
v_targets_weights <- c(1)
l_targets <- list('v_targets_names' = v_targets_names, 'Surv' = Surv,
                  'v_targets_dists' = v_targets_dists,
                  'v_targets_weights' = v_targets_weights)

v_params_names <- c("p_Mets", "p_DieMets")
v_params_dists <- c("unif", "unif")
args <- list(list(min = 0.04, max = 0.16),
             list(min = 0.04, max = 0.12))
l_params = list(v_params_names = v_params_names,
                v_params_dists = v_params_dists,
                args = args)

samples <- sample_prior_LHS(.l_params = l_params,
                            .n_samples = 10)

l_lik <- log_likelihood(.func = CRS_markov, .args = NULL,
                        .samples = samples, .l_targets = l_targets)

## End(Not run)

W-Mohammed/calibR documentation built on Oct. 16, 2023, 12:17 a.m.