The Jaffle Shop Generator or jafgen
is a python package with a simple command line tool for generating a synthetic data set suitable for analytics engineering practice or demonstrations.
If you have pipx installed, jafgen
is an ideal tool to use via pipx. You can generate data without installing anything in the local workspace using the following:
pipx run jafgen [options]
You can also install jafgen into your project or workspace, ideally in a virtual environment.
pip install jafgen
jafgen
takes two optional arguments:
--years
sets the length of time the simulation generates synthetic data for. It defaults to 2.--prefix
sets a prefix for the generated files in the format[prefix]_[file_name].csv
. It defaults toraw
.
# generate a simulation spanning 3 years from 2016-2019
jafgen --years 3
Finding a good data set to practice, learn, or teach analytics engineering with can be difficult. Most open datasets are great for machine learning -- they offer single wide tables that you can manipulate and analyze. Full, real relational databases on the other hand are generally protected by private companies. Not only that, but they're a bit too real. To get to a state that a beginner or intermediate person can understand, there needs to be an advanced amount of analytics engineering transformation applied.
Coming soon.
Coming soon.