hera

WORK IN PROGRESS

Hera is a prototype package to explore the future of regulatory environmental modelling. The package outlines possible approaches to creating a shared research platform for building, testing and deploying models used to assess environmental risk for regulatory purposes. See Request for comment paper for an outline and broad technical specifications.

Hera is envisaged as an opinionated collections of R packages designed for sharing environmental models. All packages share an underlying design, grammar and data structures. This allows the separation of the concerns between data, models, post-modelling steps and visualisation. Allowing greater collaboration and sharing of methods and tools.

Installation

Install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("ecodata1/hera")

Documentation

Read the white paper article and documentation website (work in progress)

Example

Assess some demo data for various environmental risks:

library(hera)
data <- assess(hera::demo_data)
data[1:5, c("sample_id", "parameter", "question", "response")]
#> # A tibble: 5 × 4
#>   sample_id parameter question           response                         
#>   <chr>     <chr>     <chr>              <chr>                            
#> 1 1017980   <NA>      EPSI Score TL2     97.2877525074163                 
#> 2 1017980   <NA>      EPSI Condition TL2 Minimally sedimented/unsedimented
#> 3 1101214   <NA>      EPSI Score TL2     94.6979865771812                 
#> 4 1101214   <NA>      EPSI Condition TL2 Minimally sedimented/unsedimented
#> 5 1250462   <NA>      EPSI Score TL2     97.8168378529374

Alternatively, you can view the catalogue and select which assessments to be run.

catalogue
#> # A tibble: 4 × 3
#>   assessment                data               assessment_function
#>   <chr>                     <list>             <list>             
#> 1 Macroinvertebrate Metrics <tibble [11 × 10]> <fn>               
#> 2 DARLEQ3                   <tibble [35 × 12]> <fn>               
#> 3 Bankside Consistency      <tibble [11 × 12]> <fn>               
#> 4 RICT                      <tibble [20 × 12]> <fn>

Then select which assessment(s) you wish to run by name:

assessments <- assess(demo_data, 
                      name = c("RICT",
                               "Macroinvertebrate Metrics"))
assessments[1:5, c("sample_id", "parameter", "question", "response")]
#> # A tibble: 5 × 4
#>   sample_id parameter question           response                         
#>   <chr>     <chr>     <chr>              <chr>                            
#> 1 1017980   <NA>      EPSI Score TL2     97.2877525074163                 
#> 2 1017980   <NA>      EPSI Condition TL2 Minimally sedimented/unsedimented
#> 3 1101214   <NA>      EPSI Score TL2     94.6979865771812                 
#> 4 1101214   <NA>      EPSI Condition TL2 Minimally sedimented/unsedimented
#> 5 1250462   <NA>      EPSI Score TL2     97.8168378529374