/kuanetal_varroapop_wrapper

R repository for sensitivity analyses for simulating pesticide impacts on honeybee colonies.

Primary LanguageR

DOI

beeRpop

Rwrapper for beepop/varroapop pesticide manual, private github repository

project

This is R code designed to wrap the C executable from Bob Curry/Crytal River Consulting. It consists of:

  1. the compiled C executable (Windows only) in the exe subdirectory
  2. R files that call the executable for a Monte Carlo analysis
  • 00run_neonic_varroapop.R - points to machine-specific local directories and source 01, 02, 03, 04, and 05 .R scripts
  • 01parameterize_simulation.R - modifies parameter inputs for model simulations and compiles input data into dataframe

02write_input.R - creates and saves .txt files of input flags for simulations

  • 03simulation_w_exe.R - runs executable
  • 04read_output.R - saves output results as .Rdata and saves simulation timestamps as an array locally
  • 05load_io.R - loads input and output data from local repository
  • 06sensitivity_analysis.R - calculate partial correlation coefficients, standard regression coefficients
  • 07plotting_neonic_generic.R - creates time series and sensitivity analysis figures from results dataframe
  • 07plotting_neonic_convsexp.R - comparison time series and sensitivity analysis figures of control vs exposed
  • 08plotting_errors.R - bug chasing

input

All possible customizable input flags for VarroaPop are listed in 01parameterize_simulation.R; user only needs to "release" the input flag and/or adjust distribution range by uncommenting line

output

Each VarroaPop simulation will create a .txt file of a results table when run from the command line. The text files are compiled into a 3D array in 03read_output.R and saved as .Rdata files in the output folder.

plots

  • fig_1MCproportions.pdf - proportion of simulations with selected response variables values > 0
  • fig_quantile_timeseries.pdf - timeseries plots of 1st, 2nd, and 3rd quantiles of selected response variables
  • fig_tornado.pdf - tornado diagram of partial correlation coefficient and standard regression coefficient values of selected response variables at specific timepoints