rcontroll

Ongoing

Contribution: Major contributor on package framework/ OS compatibility/ code implementation.

rcontroll integrates the individual-based and spatially-explicit TROLL model to simulate forest ecosystem and species dynamics forward in time.

rcontroll provides user-friendly functions to set up and analyse simulations with varying community compositions, ecological parameters, and climate conditions.

The TROLL simulator relies on a complex system of equations coded in C++ to efficiently simulate hundreds of individuals over hundreds of years. In contrast, the rcontroll package relies on a few functions in R to generate and provide inputs, prepare and run the simulations, and analyse the simulations through tables, figures, and maps that are easily connected to the rest of the R package ecosystem. The whole workflow can even be run for one or several simulations using a unique function (troll and stack functions). Pre-simulations functions include global parameters definition (generate_parameters function) and species and weather data input, with default values for French Guiana included in the package. Simulations are run alone (troll function) or as a stack (stack function) and stocked in corresponding trollsim or trollstack classes. The post-simulation outputs are stored in the corresponding trollsim or trollstack objects and can be accessed using object attributes (with @ in R) in the form of simple R objects, mainly data frames, or summarised and plotted with the print, summary and autoplot methods. TROLL simulation or stack objects consist of 7 attributes: (1) simulation name, (2) path to saved files, (3) simulation parameters, (4) simulation log, (5) initial and final simulation forest, (6) ecosystem metrics, and (7) species metrics. The initial and final forest of the simulation is a table containing spatialized inventories describing the individual states at the beginning and end of the simulation. The ecosystem and species metrics are two tables summarising overall and species forest trajectories with several variables, including for example net primary productivity and aboveground biomass. Simulations can be saved using a user-defined path when run and later loaded as a simple simulation (load_output function) or a stack of simulations (load_stack function). Finally, rcontroll includes a function to visualise TROLL simulations as an animated figure (autogif function) to better understand, demonstrate or teach the functioning of the TROLL model.