Accessing and plotting data from the ISIMIP Repository

These collection of notebooks include a step by step guide on how to access data directly from the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) Repository using the isimip-client library developed in Python by ISIMIP. These notebooks use the reticulate package to use the Python scripts directly in the R notebook.

Notebooks showing examples on how to create simple maps making use of the raster and ggplot2 packages are included in this repository.

Installing Python libraries

To install all Python libraries used in this repository, use the environment.yml file. You can do this from the command line, make sure you navigate to the folder containing the environment.yml file, and then type the following line of code: conda env create -f environment.yml -n CMIP6_data. The -f argument points to the file containing all libraries to be installed, while the -n contains the name to be given to the newly created environment. You can change the name of this environment if you prefer, but make sure this name is updated in all scripts when calling Python. Look for this line of code: use_condaenv("CMIP6_data") and update the name of the environment if you chose to use a different name.

You can check if the conda environment has been successfully created by typing this line of code in your command line: conda env list. The output will include the name and location of your newly created environment and it should look something similar to the following:

# conda environments:
#
base                     /file_path_to_environment/base
CMIP6_data            *  /file_path_to_environment/envs/CMIP6_data

Once we have successfully install the environment following the steps above, we will use the command line once again and type the following commands:

• conda activate CMIP6_data (if you used a different name, remember to change CMIP6_data for whatever you named your environment). This will activate the environment you created in the previous steps.
• which python (for Linux or Mac) or where python (for Windows) to locate where Python has been installed.

Open the file named .Rprofile, replace the folder path for the RETICULATE_PYTHON variable in line 2 with the Python folder path that was printed in the command line.