From 83cf1a99030ad5d3cffa6b2d59a2f1186e0e14df Mon Sep 17 00:00:00 2001 From: Stefan Krawczyk Date: Tue, 26 Jul 2022 11:40:24 -0700 Subject: [PATCH] Adds Hamilton as a development tool Hamilton is a paradigm that helps one manage a pandas code base in an opinionated manner. We have had great success with it at Stitch Fix, and it's got a growing open source following. I think it helps one to write pandas in a way that gives confidence if you want to run Pandas in production jobs. --- web/pandas/community/ecosystem.md | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/web/pandas/community/ecosystem.md b/web/pandas/community/ecosystem.md index 1d77c596c1eb0..89d243555a358 100644 --- a/web/pandas/community/ecosystem.md +++ b/web/pandas/community/ecosystem.md @@ -400,3 +400,20 @@ Learn more by reading through these issues [14468](https://github.com/pandas-dev [26766](https://github.com/pandas-dev/pandas/issues/26766), [28142](https://github.com/pandas-dev/pandas/issues/28142). See installation and usage instructions on the [github page](https://github.com/VirtusLab/pandas-stubs). + +### [Hamilton](https://github.com/stitchfix/hamilton) + +Hamilton is a declarative dataflow framework that came out of Stitch Fix. It was designed to help one manage a +Pandas code base, specifically with respect to feature engineering for machine learning models. + +It prescibes an opinionated paradigm, that ensures all code is: + +* unit testable +* integration testing friendly +* documentation friendly +* transformation logic is reuseable, as it is decoupled from the context of where it is used. +* integrateable with runtime data quality checks. + +This helps one to scale your pandas code base, at the same time, keeping maintenance costs low. + +For more information, see [documentation](https://hamilton-docs.gitbook.io/).