/AndurylPlus

A Python module for expert elicitation, specifically for use in R and Jupyter Notebook

Primary LanguageHTMLGNU General Public License v3.0GPL-3.0

Anduryl

A Python module and GUI for expert elicitation based on Cooke's classical method.

The software is based on an older application called Excalibur, and previous MATLAB and Python versions described in ANDURIL — A MATLAB toolbox for ANalysis and Decisions with UnceRtaInty: Learning from expert judgments and Update (1.1) to ANDURIL — A MATLAB toolbox for ANalysis and Decisions with UnceRtaInty: Learning from expert judgments: ANDURYL.

  • Free software: GNU license

Features

  • Set up your structured expert judgment project
  • Calculate decision makers
  • Robustness analysis
  • Visualizations
  • Work from a GUI or script/notebook

Installation

This module cannot be installed with pip or conda. To use it:

  1. Clone or download the repository;

  2. For running the stand alone GUI application, download the exe from here: https://github.com/grongen/anduryl/blob/master/pyinstaller/dist/Anduryl.exe

  3. For running the GUI from the python scripts, open an Anaconda command prompt in the github directory and run python -m anduryl.

  4. To import the anduryl module in your script or notebook, Python needs to find the anduryl module on one of the following directories:

    • the Python working directory. This can be checked in Python with:

      import os
print(os.getcwd())

    • the site-packages directory of the Python installation in the correct environment (see below). Using anaconda, this is usually C:/users/[name]/AppData/Local/Continuum/Anaconda3/Lib/site-packages

    • a user defined directory, added within the script with:

      import sys
sys.path.append('path/to/directory')

    If the downloaded directory is named "anduryl-master", rename it to "anduryl"

To get all the required dependencies working, it is advised to:

  1. Install a Anaconda Python distribution: https://www.anaconda.com/distribution/

  2. Create an environment with the required modules, by executing the following commands in an anaconda command prompt:

    conda config --add channels conda-forge
        conda config --set channel_priority strict
conda create --name [your environment name] numpy pyqt5 matplotlib

  3. Activate the created environment in an anaconda command prompt (activate [your environment name]) before running your notebook or script. A jupyter notebook or command prompt for the environment can also be launched from the Anaconda Navigator.

  4. For more information on how to use environments, see: https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html

Usage

The documentation provides a quickstart guide. The documentation can be accessed from the GUI menu under 'help'. For using the module directly from code or notebook, see the example notebook under https://htmlpreview.github.io/?https://github.com/grongen/anduryl/blob/master/notebooks/Example_Anduryl_with_scripting.html

Build

If you want to use PyInstaller to build the GUI yourself, view the readme in the pyinstaller directory.