Skip to content

Releases: NIEHS/ToxicR

1.1.2

10 Jun 18:27
3f4c44d
Compare
Choose a tag to compare
Merge pull request #12 from SciomeLLC/main

Set seed for nlopt and fix test dataset

ToxicR 1.1.1

31 Oct 08:19
bec5093
Compare
Choose a tag to compare

Cumulative Bug Fix

ToxicR v 1.1.0

25 Apr 13:35
Compare
Choose a tag to compare

This release updates the package to use the EFSA suite of continuous models.
To install:

Depending on your system, cut and paste the following code into your R terminal.

Recommended Method

Compile Yourself

If you have the package devtools, you can download and install directly from GitHub!

library(devtools)
devtools::install_github("NIEHS/ToxicR")

Note: For Windows, you will need the rtools executable available at: https://cran.r-project.org/bin/windows/Rtools/

Note: If you have a MacOS, you will need to download the GNU Scientific Library.
To do this, go to a command line and type

brew install gsl

This assumes you have HomeBrew installed. If you do not go to https://brew.sh, which will give you the instructions on how to install.

Note: For Linux, you will also need the GNU Scientific Library. The install depends on your flavor of Linux.
For Ubuntu, type

sudo apt-get install libgsl-dev

Alternative Methodology

First, install the required packages

install.packages(c("Rcpp","RcppEigen","RcppGSL","ggplot2","shiny","coda","scales","tidyverse","forcats","ggridges","doBy","multcomp","dplyr","rmarkdown", "actuar","ggpubr", "testthat","gridExtra","VIM","knitr", "modules", "plotly" ))

Windows R 4.3.0

download.file("https://github.com/NIEHS/ToxicR/releases/download/v1.10.0r2/ToxicR_23.4.1.1.0R4.3.zip", "ToxicR_23.4.1.1.0.zip")
install.packages("ToxicR_23.4.1.1.0.zip", repos = NULL, type = "win.binary")

Windows R 4.2.3

download.file("https://github.com/NIEHS/ToxicR/releases/download/v1.10.0r2/ToxicR_23.4.1.1.0R4.2.3.zip", "ToxicR_23.4.1.1.0.zip")
install.packages("ToxicR_23.4.1.1.0.zip", repos = NULL, type = "win.binary")

MacOS R 4.3 (M1)

download.file("https://github.com/NIEHS/ToxicR/releases/download/v1.10.0r2/ToxicR_23.4.1.1.0.tgz", "ToxicR_23.4.1.1.0.tgz")
install.packages("ToxicR_23.4.1.1.0.tgz", repos = NULL, type = "mac.binary")

23.4.1.1.0

24 Apr 13:49
378e4e6
Compare
Choose a tag to compare

This release updates the package to use the EFSA suite of continuous models.
To install:

Depending on your system, cut and paste the following code into your R terminal.

Recommended Method

Compile Yourself

If you have the package devtools, you can download and install directly from GitHub!

library(devtools)
devtools::install_github("NIEHS/ToxicR")

Note: For Windows, you will need the rtools executable available at: https://cran.r-project.org/bin/windows/Rtools/

Note: If you have a MacOS, you will need to download the GNU Scientific Library.
To do this, go to a command line and type

brew install gsl

This assumes you have HomeBrew installed. If you do not go to https://brew.sh, which will give you the instructions on how to install.

Note: For Linux, you will also need the GNU Scientific Library. The install depends on your flavor of Linux.
For Ubuntu, type

sudo apt-get install libgsl-dev

Alternative Methodology

First, install the required packages

install.packages(c("Rcpp","RcppEigen","RcppGSL","ggplot2","shiny","coda","scales","tidyverse","forcats","ggridges","doBy","multcomp","dplyr","rmarkdown", "actuar","ggpubr", "testthat","gridExtra","VIM","knitr", "modules", "plotly" ))

Windows R 4.3.0

download.file("https://github.com/NIEHS/ToxicR/releases/download/v1.10.0/ToxicR_23.4.1.1.0.R4.3.zip", "ToxicR_23.4.1.1.0.zip")
install.packages("ToxicR_23.4.1.1.0.zip", repos = NULL, type = "win.binary")

Windows R 4.2.3

download.file("https://github.com/NIEHS/ToxicR/releases/download/v1.10.0/ToxicR_23.4.1.1.0R4.2.3.zip", "ToxicR_23.4.1.1.0.zip")
install.packages("ToxicR_23.4.1.1.0.zip", repos = NULL, type = "win.binary")

MacOS R 4.3 (M1)

download.file("https://github.com/NIEHS/ToxicR/releases/download/v1.10.0/ToxicR_23.4.1.1.0.tgz", "ToxicR_23.4.1.1.0.tgz")
install.packages("ToxicR_23.4.1.1.0.tgz", repos = NULL, type = "mac.binary")

Cumulative Fixes

08 Nov 12:33
Compare
Choose a tag to compare

There have been a number of deprecated C++ functions that flagged warnings on CRAN. This is a cumulative update changing those function calls. There is no new functionality since 22.8.1.0.2.

ToxicR 22.8.1.02

02 Aug 00:46
Compare
Choose a tag to compare

Version 22.8.1.0.2

The following fixes are in version 1.0.2:

- Function 'single_continuous_fit' and 'ma_continuous_fit' changed error when defining default priors
	 for 'distribution=normal-ncv' when data are negative. Originally the variance was described as mean(Y)/var(Y); 
	however, for negative means, this causes NA error. It is now defined as abs(mean(Y))/var(Y). 
- Log-normal distribution fits were incorrect when summarized data was used. The correct transformation of
	summarized data is now performed. The formula for standard deviation was typed in as sqrt(log((sd)^2/mu + 1)) it is now sqrt(log((sd/mu)^2+1)). 
- Changed default priors for dichotomous fits to be consistant with Wheeler et al. (2020). 

The following changes to fitting were made:

- Changed MLE Polynomial fit behavior.  Now the terms up to the quadratic are constrained to be in the direction 
  of the response.  After this, i.e., degree >= 3, the parameters are unconstrained. 
- Added summary and print methods for mcmc model averaging. 

Known Problems not yet fixed

- GoF for MA individual models not given. 
- GoF for dichotomous models with (0/1) data fails. 

ToxicR 22.01 (1.0.0)

28 Jan 17:44
Compare
Choose a tag to compare

Initial Release of ToxicR 1/28/2022.

Executable files are for R 4.1.2 in Windows/Mac(Intel/M1).

  • The macOS release is single threaded due to OpenMP issues.
  • The Windows release is multi-threaded
  • For Linux compiles GSL (2.6 or greater) and NLOPT (2.4 or greater) libraries need to be installed.

Release Notes:

Dichotomous and Continuous Dose-Response functionality

Dichotomous

  • MLE functionality (Equivalent to EPA BMDS 3.x)
  • Bayesian MAP/Laplace (Equivalent to EPA BMDS 3.2+)
  • Bayesian MCMC estimates
  • Bayesian Model Averaging (MAP Equivalent to EPA BMDS 3.2+)

Continuous

  • MLE functionality (Equivalent to EPA BMDS 3.x)
  • Bayesian MAP/Laplace
  • Bayesian MCMC estimates
  • Bayesian Model Averaging

NTP Bioassay Tests

  • Dose Dependent Trend Tests (e.g. Williams etc)
  • PolyK Test

Methodologies Described in
Bailer, A.J. and Portier, C.J., 1988. Effects of treatment-induced mortality and tumor-induced mortality on tests for carcinogenicity in small
samples. Biometrics, pp.417-431.

Wheeler, M.W., Blessinger, T., Shao, K., Allen, B.C., Olszyk, L., Davis, J.A. and Gift, J.S., 2020. Quantitative Risk Assessment:
Developing a Bayesian Approach to Dichotomous Dose–Response Uncertainty. Risk Analysis, 40(9), pp.1706-1722.

Wheeler, M.W. Cortinas,J. Aerts, M. Gift, J.S. Davis J.A., 2022. Continuous Model Averaging for Benchmark Dose Analysis: Averaging Over
Distributional Forms. Resubmitted to Environmetrics.