Skip to main content

Interfaces for an in-memory datastore and calc framework using marshmallow + pandas

Project description

binx Documentation Status


binx is a small Python framework for application data modeling and transformation. It’s API relies heavily on marshmallow for validation, object serialization and storage. It’s true purpose however is to expose an API that allows developers to model procedural code into directed, acyclic graphs of arbitrary complexity. A user can define an application or library as a family of binx.Collection objects (nodes) and use the binx.adapter module (edges) to create a network of data transformations while guaranteeing data integrity along the way.

The main goal of the project is to provide a simple API for data scientists, engineers or developers who write alot of procedural code to be able to organize their projects using a declarative style similar to how one might approach writing a web application using DRY principles.



This set of interfaces are designed to help you quickly scale up your notebooks/scripts and create uniformity between your projects!

binx provides:

  • A declarative style in memory datastore (

  • A declarative Adapter API that helps model/manage relationships and data transformations between collections (

  • consistent API for moving your data between json, py-objs, and pandas dataframes


0.1.2 (2018-05-28)

  • Port initial API from bema project

  • Port original unittests

  • change marshmallow api to 2.x

0.2.0 (2018-07-03)

  • Built adapter module and related functionality

  • removed calc_factory API (possibly will be re-included in a later build)

  • created the CollectionBuilder class

0.2.1 (2018-07-12)

  • patched issue relating to datetime/datestrings not being parsed by collections

0.2.2 (2018-07-26)

  • patched issue relating to introspection of required columns (issue #7-#8)

0.2.3 (2018-10-01)

  • clean up and fixes to cli

0.3.0 (2019-07-14)

Fixes some long standing issues and adds some new features.

  • - kwarg to accumulate optionally accumulate intermediate collections in adapter chain

  • - A user warning is issued instead of exception if an identical class path name is overwritten

  • - fixed errors related to creating dataframes from NoneType

  • - added a new base class PluggableAdapter to ease development of the adapter chain

  • - now excepts an optional name arg. Will attempt to auto-parse name from serializer_class.

0.3.1 (2019-07-15)

  • Adds a context variable to the AdapterChainError that can be accessed if the error is caught

0.3.2 (2019-08-21)

Final release that is compatible with marshmallow 2.x

  • Pins marshmallow to v0.20.2

  • Final release supporting marshmallow 2.x functionality

0.4.0 (2019-08-23)

A minor release that internally breaks backward compatibility with marshmallow-2.x. Adds new feature to that allows loading data and setting ma configuration directly from Collection constructor.

  • private api changes to Serializer and Collection classes.

  • data can now be loaded via the Collection constructor.

  • kwargs can be passed to the underlying ma schema at runtime.

0.4.1 (2019-09-23)

A patch that pins the library to marshmallow-v3.1.1

0.4.2 (2019-10-27)

Bugfix that fixes the dependency in the previous patch. This version is compatible with marshmallow>=3.1.1. Also assures pandas>=0.23 in

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

binx-0.4.2.tar.gz (36.3 kB view hashes)

Uploaded source

Built Distribution

binx-0.4.2-py2.py3-none-any.whl (16.1 kB view hashes)

Uploaded py2 py3

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Fastly Fastly CDN Google Google Object Storage and Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page