AI for Illicit Trade

Directory structure

Peregrine
|____
     |__ illicitAI (Docker container & Git repo)
VM   |            |__ illicitAI.Rproj
|____|            |__ Data (run postBuild.sh or download from Zenodo)
                  |   |__ CEPII
                  |   |__ FATF
                  |   |__ IFF
                  |   |__ IMF
                  |   |__ TI
                  |   |__ TJN
                  |   |__ WITS
                  |   |__ WB
                  |   |__ WB
                  |__ Figures (generated by the scripts)
                  |__ Results (generated by the scripts)
                  |__ Scripts
                  |   |__ Data Preparation.R
                  |   |__ Analysis.R
                  |__ docker-compose.yml
                  |__ Dockerfile
                  |__ postBuild.sh
                  |__ requirements.R
                  |__ setup.sh

Run the code live in Jupyter Lab     

Run the code live in RStudio     

Set up

You can also clone this repo on your machine and spin it up in a Docker container. Requirements: docker.

git clone https://github.com/walice/illicitAI.git
cd illicitAI
docker-compose up --build -d

Then, run the postBuild.sh script to download the Data.

chmod +x postBuild.sh
./postBuild.sh

Run the makefile.

Rscript ~/work/Scripts/DataPreparation.R
Rscript ~/work/Scripts/DataPreprocessing.R
Rscript ~/work/Scripts/Analysis.R

Execute the notebooks via papermill.

jupyter run execute-notebook.py