Bayesian Inference of TKTD Models

Advanced methods for a valuable quantitative environmental risk assessment using Bayesian inference of survival Data with toxicokinetics toxicodynamics (TKTD) models. Among others, it facilitates Bayesian inference of the general unified threshold model of survival (GUTS). See models description in Jager et al. (2011) and implementation using Bayesian inference in Baudrot and Charles (2019) .

morseTKTD

Advanced methods for a valuable quantitative environmental risk assessment using Bayesian inference of Survival TKTD models (like the Generalized Unified Threshold model of Survival (GUTS)).

Submission

Before a submission, you can look at prepare-for-cran , which is an open and collaborative list of things you have to check before submitting your package to the CRAN.

Otherwise, check "as-cran"" using the source package:

library(devtools)
# create documentation
devtools::document(roclets = c('rd', 'collate', 'namespace'))

Once the archive is done, check that .Rbuildignore was applied. Try to have a low size archive (< 2Mb)

Either directly

# build and check the archive
devtools::check()

Or in 2 steps:

# 1. build the package. 
devtools::build()
# 2. check the archive. 
devtools::check_built("../morseTKTD_0.1.0.tar.gz")

See the CRAN status of your sumbmission:

• incoming R CRAN packages: Index of /incoming
• incoming dashboard: incoming dashboard

Install from gitlab repository

library('remotes')
remotes::install_gitlab("mosaic-software/morsetktd", host = "gitlab.in2p3.fr")

Build the manual and vignettes

library('devtools')
devtools::document(roclets = c('rd', 'collate', 'namespace'))
devtools::build_manual()
devtools::build_vignettes()

building the package

Note add to .buildignore

# remove files .rds in fixtures
rds_files <- list.files(path = "tests/testthat/fixtures", pattern = "\.rds$", full.names = TRUE)
use_build_ignore(rds_files, escape = TRUE)

library(devtools)
devtools::build()

Add dependencies

usethis::use_package("ggplot2")

Coverage

From R session

library(covr)
cov <- package_coverage("morseTKTD")

Style of process

The succession of steps

1. data: load the data set.
2. BinaryData, CountData or ContinuousData: make a ModelData object for binary, count and quantitative continuous data, respectively.
3. The above-mentioned objects inherit of data.frame
4. plot: plot a ModelData object.
5. summary: provides a summary of a ModelData object.
6. doseResponse: return a DoseResponse object.
7. plot: plot a DoseResponse object.
8. fit: fit a ModelData object and return a Fit object.
9. plot: plot a Fit object.
10. ppc: return a PPC object.
11. plot: plot a PPC object.

Coding Style

Object: BigCamelCase

class(x) <- append("ObjectCamelCase", class(x))

Methods: small_snake_case

methods_snake_case <- function(object, ...){
  UseMethod("methods_snake_case")
}

methods_snake_case.ObjectCamelCase <- function(...){}

Function (no methods - not linked to object): smallCamelCase

smallCamelCase <- function(...){}