nwb-project-analytics

Repository for collecting analytics and scripts related to the NWB project.

The NWB project maintains and contributes to a large number of codes related to NWB. The goal of this effort is to help developers to get a quick overview of the state of NWB code repositories. See, e.g., the code health page

Status: This project is under active development. The code is in the alpha development phase.

How to use nwb-project-analytics (the docs)

All main NWB analytics are compiled automatically when building the docs

1. Create a virtual environment (option)

conda create --name nwb_analytics_env python=3.8
conda activate nwb_analytics_env

2. Install cloc (see here) (only needed when you want to update the cloc code statistics in data/)
3. Install the tools via

git clone https://github.com/NeurodataWithoutBorders/nwb-project-analytics.git
cd nwb-project-analytics
pip install -r requirements.txt -r requirements-doc.txt
pip install -e .

4. Build the docs

cd docs
make html
open build/html/index.html

How to force rebuild of all figures and apidoc

cd docs
make allclean

Using the make allclean command removes all auto-generated figures and rst files in the docs/source directory (specifically docs/source/code_stat_pages and docs/source/nwb_project_analytics.*rst) as well as all builds from docs/builds. When rebuilding the docs (e.g., via make html) the files will be regenerated (using the data cached in data/).

How to manually update the data/

On this repo, the information cached in data/ is automatically updated nighlty by the build_analytics_data.yml GitHub Action. To manually update the statistics cached in data/, simply remove the cached results and build the docs:

rm data/cloc_stats.yaml
rm data/commit_stats.yaml
rm data/git_paths.yaml
rm data/release_timelines.yaml
rm data/contributors.tsv
cd docs
make html

After completing the update, commit the updated data to the repo

git commit -m "Updated code statistics data" ./data/*.yaml

How to add a new code

To add a new entry to the NWBGitInfo.GIT_REPOS dictionary in src/nwb_project_analytics/gitstats.py. The dictionary is used to track all main NWB repositories and stores for each repo a `` GitRepo`` object with basic metadata about the code (e.g., the location of the repo, name of the main branch, etc.). When adding a new code, all statistics need to be updated following the instructions above on How to update the ``data/``.

How are code statistcs computed?

The code uses the cloc tool to calculate line-of-code statistics. Computing cloc statistics for all codes is time-consuming as we need to compute them over time. This is accomplished by running cloc for all commits (or more accurately the last commit on each day) on the main branch of each code. The cloc results are saved in the data/ folder to avoid unnecessary updates and safe time.

How to build custom analytics

The nwb_project_analytics library includes a number of tools to help with collecting and plotting data related to NWB code repositories. This includes for example:

• nwb_project_analytics.codecovstats : Module for getting data from Codecov.io
• nwb_project_analytics.codestats : Module for computing code statistics using CLOC
• nwb_project_analytics.gitstats : Module to help query GitHub repos
• nwb_project_analytics.renderstats module : Module for creating plots for code statistics
• nwb_project_analytics.create_codestat_pages : Module used to generate Sphinx documentation with code statistics