A use for One-At-a-Time sensitivity analysis: Diagnostic testing of integrated models

Companion code for the paper detailing the analysis approach and figure generation.

Setup

Clone the necessary repositories:

$ git clone https://github.com/ConnectedSystems/oat-use.git
$ git clone https://github.com/ConnectedSystems/SALib.git --branch radial-oat-method --single-branch salib-roat
$ cd salib-roat
$ git checkout da99fdaed05c29e98ab8d7685d2c6ad3783ce946
$ cd ..

Set up environment from the project folder:

$ cd oat-use
$ conda create -n oat-use python=3.6.6 -y
$ conda activate oat-use
$ pip install -r requirements.txt
$ cd ..

Install the specific branch of SALib used for the study:

$ cd salib-roat
$ pip install -e .
$ cd ..

Set up ipykernel with:

$ python -m ipykernel install --name oat-use --display-name "Python (oat-use)"

The full instructions as above:

$ git clone https://github.com/ConnectedSystems/oat-use.git
$ git clone https://github.com/ConnectedSystems/SALib.git --branch radial-oat-method --single-branch salib-roat 
$ cd salib-roat
$ git checkout da99fdaed05c29e98ab8d7685d2c6ad3783ce946
$ cd ..

$ conda create -n oat-use python=3.6.6 -y
$ conda activate oat-use
$ cd oat-use
$ pip install -r requirements.txt
$ cd ..
$ cd salib-roat
$ pip install -e .
$ cd ..
$ python -m ipykernel install --name oat-use --display-name "Python (oat-use)"

Notebooks

Notebooks are found in the notebooks directory and are labelled in order of use.

Miscellaneous notebooks are unnumbered and are included for posterity.

To view notebooks locally, it is assumed jupyter notebook or lab is installed.

pip install jupyterlab

Once the conda environment is activated:

$ cd notebooks
$ jupyter lab

Scripts

Scripts used to generate samples are found in the scripts directory. These are to retain a record of sample provenance and do not need to be run.