GCRMN model development

Building the environment

Most of the dependencies can be inferred by examination of the Dockerfile. In fact, the safest way of ensuring that the codes will run is the build a docker image from the Dockerfile and run within a container.

Running in the container

To run these models inside the docker container (after building the container), use the following steps:

1. make R_container: this will run the the following analyses in R:

• synthetic_data(): to generate the synthetic data
• site_replacement(): R based models for exploring site replacement scenarios
• missing_years(): R based models for exploring missing years scenarios
• incomplete_spatial(): R based models for exploring incomplete spatial scenarios

2. make python_container: this will run the the following analyses in R:

• site_replacement(): python based models for exploring site replacement scenarios
• missing_years(): python based models for exploring missing years scenarios
• incomplete_spatial(): python based models for exploring incomplete spatial scenarios

3. make R_container: a second time (after make python_container) in order to combine the previous R and python models together for the purpose of comparing the models
4. make docs_container: this will render each of the quarto documents

Running the codes

Nevertheless, if you are running on bare metal, then ensure that both R and python are installed and that their respective packages indicated in the Dockerfile are installed and available.

Ideally, it should be possible to run in the following order:

1. make run_R: this will run all the R based analyses using the targets package to ensure all steps are performed in the correct order
2. make run_python: this will run all the python based analyes using the ploomber library to ensure all steps are performed in the correct order
3. make run_R: a second time (after make run_python) in order to combine the previous R and python models together for the purpose of comparing the models
4. make render_docs: this will render each of the quarto documents

I say ideally, because in my haste and to quickly knit together the python analyses and R analyses so that they can be compared, I included R code to read in the results of the python analyses. Of course the first time that this all gets run, those python analyses wont occur and so it will error. Unfortunately, it is not as simple as just running all the python analyses first (e.g. swapping steps 1 and 2 above) because the python analyses rely on the simulated landscapes that are created early on in the R analyses.

I should re-write this to be a bit more granular so that it runs the first part of the R analyses, then python, then the later R steps - however I ran out of time! I will hopefully do this step soon.

Github actions to publish the documents as github pages

I have written github workflows to automate the whole process on a git commit. Unfortunately, the workflow hits the hard walltime of 6hrs and therefore fails. I am currently exploring options for engaging a runner to allow full continuous integration.