Skip to content

Latest commit

 

History

History
76 lines (48 loc) · 3.2 KB

README.md

File metadata and controls

76 lines (48 loc) · 3.2 KB

What are these scripts for?

How to use these scripts

The scripts beginning with 1.x, 2 and 3 are the main scripts to be used. The 1.x scripts are the three scripts for pre-processing all the data and the 2 scripts are for running prioritization. The 3 scripts are for analyzing the outputs.

0.1-ramsar.R contains some separate processing for ramsar sites. 0.9-helper_functions.R contains the functions used by 1.2 and 1.3

See below for basic descriptions of all the scripts, particularly the analysis ones.

Pre-processing

This stage has three scripts:

  • 1.1-OPTIONS.R -- this sets the shared options for all the scripts in 1.2, 1.3 and 2x so that they aren't repeated.
  • 1.2-MAIN_PREPROCESS_master.R -- this is the primary script for pre-processing and should run almost completely automatically apart from a couple of lines at the beginning to be edited
  • 1.3-split_by_country_optional.R -- this is used if you want to run prioritizr with a separate feature for each country

Prioritization

This stage only needs one script, but there are multiple versions for different uses:

  • 2a-prioritzr_autoExc_master.R -- main multi-budget spatial prioritization script with easy exclusion of features.

  • 2b-prioritzr_single_master.R – run prioritization for a single budget

Archived scripts that need to be updated

  • 2c-prioritzr_per_country.R – run prioritization for each country separately (not globally) using the same preprocessing as 1.x

  • 2d-prioritzr_sensitivity.R – run a single budget prioritization multiple times to compare times and solutions (though noting that past tests have led to the same solution every time if all input is the same).

Analysis

There are a number of scripts to aid in different parts of analyzing the outputs of the pre-processing and the solutions. They rely on 1.1-OPTIONS.R and 3.0-helper_functions.R.

General analysis:

  • 3-exclusion_area.R - create a raster with the reason for exclusion for the exclusion raster.
  • 3-num_features.R - create raster with the number of features that are available in each point
  • 3-stats.R - create some stats for restorable land
  • 3-explore_processed.R - compare the pre-processed outputs from different pre-processing

Solution analysis

  • 3-es_values.R - calculate and plot cumulative amount of ecosystem service by rank
  • 3-sample.R - take a random sample of the solution and run correlation and regression
  • 3-explore_incomplete_solution.R - combine and rasterize partial solutions (i.e., not all budgets) for analysis
  • 3-compare_runs.R - compare solutions from different types of runs (i.e., different runids)

Other:

  • 3.0-helper_functions.R - a collection of helper functions for the analysis

Script package dependencies

The following R packages are required:

  • tidyverse
  • sf
  • terra
  • arrow
  • data.table
  • prioritizr
  • rcbc or lpsymphony
  • tictoc
  • glue

Needed updates/future features

Things that need fixing

  • Ensure that the directories for output and logs are empty before writing (to catch not changing run_id)

Enhancements

  • Check that everything uses the extent
  • Force checking/plotting outputs at the end