Gemd

What’s New
* The cake examples has been extended to demonstrate more of the data model
* The Strehlow and Cook example now imports directly from the Citrination PIF and includes a substantially increased number of attributes

What’s New

* As our experience with the data model grows, we are working hard to both improve the tools for interacting with it and pushing its structure forward, making to more representative of the real world information Materials Scientists work with every day. This release represents both improvements in those interfaces as well as fixing some bugs.

Improvements

* The serialization and deserialization as well as the helper functions to interact with and modify GEMD graphs have been improved. This should empower users to get more done in fewer lines of code.
* The data example Cake has been augmented to have complete coverage of the data model. This should help users ground their understanding of Templates, Objects and Values in a practical, relatable resource.

Fixes

* IngredientRuns and IngredientSpecs have been modified so that the logical equivalence of their name and labels fields is easier to have consistent when objects are being created.
* And, of course, all of this is tested and documented

Coming Soon

* More examples. We understand how much easier it is to understand a data model with concrete examples, so we're working hard to develop more and to use those examples in end-to-end demonstrations of how to use our entire data management and machine learning stack.

The largest change with this release is a large increase in the list of supported units. We've also improved documentation, particularly for the installation process in our README.

What's new:

* Augmented unit sets
* Improved documentation of installation, testing and units coverage
* Updated keyword handling for serialization technology
* Augmented test coverage

Coming Soon:

* More types for supporting data provenance

Most of this release is test coverage, which includes some changes to the exception types to better reflect the error that is being raised. This release also includes improvements to the string representation of data objects.

What's new
* 100% test coverage.
* A more intelligent `__repr__` method, which will include the names of soft-linked objects.

Changes
* Some `ValueError` changed to `TypeError` when validating arguments. This better reflects the error mode.

Coming Soon
* More units for real values, e.g. 23

This is the first beta release of `taurus-citrine`: the python binding to Citrine's next data format, also code-named [taurus](https://citrineinformatics.github.io/taurus-documentation/). It supports access to the Citrine Data Platform via [citrine-python](https://github.com/CitrineInformatics/citrine-python). It is a beta release that we expect to actively develop over the next several months.

What's new
* Everything! This is the first release of `taurus` and the first implementation of the `taurus` data model.
* Material histories with Materials, Processes, Ingredients, and Measurements
* Differentiating between intent and reality with Specs and Runs
* Opt-in data validation with Templates
* Opt-out uncertainty quantification with distribution types
* Multi-object serialization and deserialization with link normalization
* Unit string parsing

Coming Soon
* More units for real values, e.g. 23
* More sources for `MeasurementRun` and `ProcessRun`