This repository contains a set of Jupyter notebooks which develop summary statistics and charts for visualizing the comparative results of a base and build scenario. The initial use case is a bespoke solution for the Explicit Error Terms project but may be useful for any base-vs.-build scenario comparison.
The notebooks, stored in the notebooks directory, all
pull base and build data from directories specified in
_quarto.yml under the sources setting. An example
dataset pair is provided under the input/example directory
which acts as the default source in the configuration YAML.
Modify the sources.base and sources.build settings to
select a different data source.
To install the appropriate dependencies for creating the
output visualizations, install the dependencies in the
environment.yml file using conda env create -n {environment_name}
To build the output website, run the compile.py script, which
will pull a list of Jupyter notebooks from _quarto.yml to
execute. The outputs from these notebooks will be compiled
into the new website, found in the output directory.
When compiling the source notebooks, output data and metadata
are generated and stored in the file alongside the source.
In preparation for uploading notebooks to this repository, please remove the output data and metadata to keep the notebooks tidy.
Functionality is included in the compile.py script to do this
using the --clean command-line argument. For example:
python compile.py --clean