Skip to contents

Introduction

Source: vignettes/Introduction.Rmd
Introduction.Rmd

This R package is designed to facilitate robust statistical analyses, particularly in the realm of dose-response modeling. Our package provides a suite of tools and examples that enable users to analyze datasets effectively, leveraging various statistical methods to draw meaningful conclusions. Pros and cons for different methods are illustrated. Verification or validation of certain functions from certain packages are also included.

library(drcHelper)
#> Loading required package: drc
#> Loading required package: MASS
#> Loading required package: drcData
#> 
#> 'drc' has been loaded.
#> Please cite R and 'drc' if used for a publication,
#> for references type 'citation()' and 'citation('drc')'.
#> 
#> Attaching package: 'drc'
#> The following objects are masked from 'package:stats':
#> 
#>     gaussian, getInitial

The website is built using the pkgdown framework. There are 4 sections other than the function references page: Get started, Examples, Articles, and Validation.

Getting Started

In the Getting Started section, we introduce the main helper functions that serve as the foundation for your analyses. Here, you’ll find typical workflows that illustrate how to navigate through the functionalities of our package. It’s important to note that there are multiple approaches to analyzing a dataset, and the choice of method can significantly influence the results. When the dataset contains information with a degree of certainty, different analytical methods should converge on similar conclusions. Conversely, when the data lacks sufficient information to address a specific question, disparate methods may yield distinct and potentially conflicting results. Thus, characterizing uncertainty is a crucial component of any statistical analysis, and our package emphasizes this aspect throughout its functionalities.

For routine analyses, we adopt the drc package as our core analysis tool, given its robust capabilities for dose-response modeling. However, we recognize that there are numerous other packages designed for similar analyses, such as drda, DoseFinding, and others. In the Articles section, we provide examples and reproduce analyses using these alternative packages, allowing users to explore different methodologies and gain insights into how similar analyses can be conducted across various frameworks. Additionally, for specific types of data, we offer supplementary analyses utilizing other testing or modeling approaches, ensuring a well-rounded perspective on statistical evaluation.

Regulatory Stats

The Regulatory Stats section explains some regulatory concepts and showcases practical applications of our package’s functionalities, demonstrating how to implement analyses in real-world scenarios. These examples serve as a guide for users to replicate analyses and adapt them to their own datasets.

Articles

In the Articles section, we delve deeper into various topics related to dose-response analysis, providing insights, comparisons, and discussions on methodologies. This section aims to broaden your understanding of statistical approaches and their implications in research.

Validation

Lastly, the Validation section outlines the validation processes employed within our package. Here, we discuss how we ensure the reliability and accuracy of the methods implemented, providing users with confidence in the results generated by our analyses.

Other useful packages

There are several other packages I find useful for the statistical analysis of (eco)toxicological data.

  • ecotoxicology package

    • Implementation of the EPA’s Ecological Exposure Research Division (EERD) tools (discontinued in 1999) for Probit and Trimmed Spearman-Karber Analysis.
    • Probit and all the tables from Finney’s book (code-generated, not copied) with the generating functions included.
    • Control correction: Abbott, Schneider-Orelli, Henderson-Tilton, Sun-Shepard. Toxicity scales: Horsfall-Barratt, Archer, Gauhl-Stover, Fullerton-Olsen, etc.

  • drda - Has Gompertz and log-gompertz function - for both growth curves and dose-response analysis - Probably with better optimiation algorithms compared to drc.

  • bayesnec