rueedlinger / monte-carlo-simulation

Monte Carlo Simulation for Forecasting Cycle Times with a Python Jupyter Notebook

Monte Carlo Simulation for Forecasting Cycle Times

I came across some examples and plugins (see references) for estimating the cycle times in a Kanban system. So my goal was to build the same thing with a Python Jupyter Notebook.

The parts:

• The Data Exploration Notebook (notebooks/analyse_data.ipynb) can be used to examine the data.
• The Forecasting Notebook (notebooks/forecasting_with_monte_carlo.ipynb) is used to do a Monte Carlo simulation for the cycle times.
• The data as CSV file.
  • id => the JIRA issue id (only the number part).
  • grp => the project part from the JIRA issue as number
  • cycle_time_days => cycle time in days for this JIRA issue (finish date - start date = cycle time days).
  • created_date => when the issue was created

Getting Started

All the required Python packages can be installed with pipenv.

Project Setup

First you nee to install pipenv.

$ pip install --user pipenv

Install all the required packages

$ pipenv install --dev

Run the Notebook

pipenv run jupyter-lab

CI Build

References

• https://www.actionableagile.com/
• https://kanbanize.com/blog/monte-carlo-simulations/