From bef1409236082b5194f7d1f184b7afdd1a11a657 Mon Sep 17 00:00:00 2001 From: Nils Kehrein Date: Mon, 23 Sep 2024 09:22:15 +0200 Subject: [PATCH] release of v1.2.0, minor doc update --- NEWS.md | 2 +- R/lik_profile.R | 7 +++++-- man/lik_profile.Rd | 11 +++++------ 3 files changed, 11 insertions(+), 9 deletions(-) diff --git a/NEWS.md b/NEWS.md index b66af31..2fb669f 100644 --- a/NEWS.md +++ b/NEWS.md @@ -1,4 +1,4 @@ -# cvasi Development version +# cvasi 1.2.0 * New likelihood profiling with `lik_profile()` based on the routines described by Tjalling Jager (doi: 10.1002/ieam.4333) and implemented in *BYOM* diff --git a/R/lik_profile.R b/R/lik_profile.R index 7f3dc09..d3db30b 100644 --- a/R/lik_profile.R +++ b/R/lik_profile.R @@ -1,9 +1,12 @@ #' Likelihood profiling #' -#' @description The aim of the function is 2-fold: 1) estimate a 95% confidence +#' @description +#' `r lifecycle::badge("experimental")` +#' +#' The aim of the function is two-fold: 1) estimate a 95% confidence #' around each parameter of a calibrated model, and 2) see if perhaps a local minimum was found rather #' than a global minimum. To achieve this, the likelihood profiling goes through -#' every parameter one by one (using [profile_par())]. For each parameter, +#' every parameter one by one. For each parameter, #' the model is sequentially refit with the parameter value set to #' increasingly lower and higher values, and the likelihood of the model given the #' data calculated (using [log_lik()]). The likelihood is then compared diff --git a/man/lik_profile.Rd b/man/lik_profile.Rd index ec645eb..d90b165 100644 --- a/man/lik_profile.Rd +++ b/man/lik_profile.Rd @@ -44,13 +44,15 @@ the 95\% confidence interval; the original parameter value; the likelihood plot the recalibrated parameter values (in case a lower optimum was found) } \description{ -The aim of the function is 2-fold: 1) estimate a 95\% confidence +\ifelse{html}{\href{https://lifecycle.r-lib.org/articles/stages.html#experimental}{\figure{lifecycle-experimental.svg}{options: alt='[Experimental]'}}}{\strong{[Experimental]}} + +The aim of the function is two-fold: 1) estimate a 95\% confidence around each parameter of a calibrated model, and 2) see if perhaps a local minimum was found rather than a global minimum. To achieve this, the likelihood profiling goes through -every parameter one by one (using [profile_par())]. For each parameter, +every parameter one by one. For each parameter, the model is sequentially refit with the parameter value set to increasingly lower and higher values, and the likelihood of the model given the -data calculated (using [log_lik()]). The likelihood is then compared +data calculated (using \code{\link[=log_lik]{log_lik()}}). The likelihood is then compared to the likelihood of the original model (using a likelihood ratio). This leads to the development of a likelihood profile, from which a plot a 95\% confidence interval for the parameter is derived. @@ -70,9 +72,6 @@ something small). The function was inspired by a MatLab BYOM v.6.8 procedure, created by Tjalling Jager. For details, please refer to BYOM (http://debtox.info/byom.html) as well as Jager (2021). - -[profile_par())]: R:profile_par()) -[log_lik()]: R:log_lik() } \examples{ # Example with Lemna model - physiological params