Citrine

We are excited to announce some highly impactful new features with this release of Citrine Python. You can now get the fully featurized training data associated with any valid predictor. We've also included more control over the generative molecular design process, as well as improved the data ingestion process.

What's New
* Finally, you can quickly access the fully featurized training data from a given predictor with the `get_featurized_training_data` method. Simply call `project.get_featurized_training_data` with the ID and version of your predictor and you will get a list of design materials representing your training data with features present. 858
* You can now designate the minimum number of substructures to construct when running generative molecular design by passing a dictionary to the new `min_substructure_counts` argument. 856
* Thanks to 859, a slew of new arguments to the ingest method make getting structured data on platform substantially easier:
* `upload` will copy files onto platform if they're not there already
* `delete_dataset_contents` and `delete_templates` will clear out old records prior to creating new ones
* `build_table` will follow object generation with creating a table for that data, ready for modeling

Improvements
* FileLink objects and the associated FileCollection methods (read, write, etc.) now work smoothly with local files just like remote ones: just use the [File URI scheme](https://en.wikipedia.org/wiki/File_URI_scheme) or convenient `from_path` method. #859

**Full Changelog**: https://github.com/CitrineInformatics/citrine-python/compare/v2.19.2...v2.22.0

Maintenance only release. Nothing exciting here.

Fixes
* Fixed an out-dated url in the ingest method 854
* Updated the python version used in the generation of the citrine-python documentation 855

**Full Changelog**: https://github.com/CitrineInformatics/citrine-python/compare/v2.19.0...v2.19.2

A handy new option and a little more maintenance.

What's New
* We've added an option when deleting the contents of a dataset to skip templates. This means that, for a pesky corpus that needs to be rebuilt rather than just updated, your templates can persist, and thus all your tables and AI artifacts will still work. 853

Deprecated
* We've deprecated the code associated with some design space reuse approaches the platform no longer supports 852


**Full Changelog**: https://github.com/CitrineInformatics/citrine-python/compare/v2.17.1...v2.19.0

One new feature for generative design, and a whole bunch of maintenance work.

What's New
* Structure exclusion! As you mutate your base molecules to explore new chemistries, you can now specify what types of groups you don't want to consider. 848

Improvements
* We've reworked the relationships between our various libraries, which should result in faster code in your Jupyter notebooks and ETL scripts.
* And a better safety net: we've added Python 3.11 to our regular testing protocol. 851

**Full Changelog**: https://github.com/CitrineInformatics/citrine-python/compare/v2.15.0...v2.17.1

Bug fixes, a bit of refinement, and more bug fixes. No new features for this release, but several corrections to make our interfaces work more smoothly.

Improvements
* We've simplified the predictor classes to better reflect how the Platform operates. All Predictors are Graph Predictors and we've added a new class to represent the nodes of that graph. This should reduce the coding effort for defining and interacting with your models. 844
* We've added a method to the Predictor Collection for asynchronously making a default predictor, if that's more your style. 849

Fixes
* We've corrected some issues with how the SDK was defining ingredient ratio constraints 845
* We've made some unsupported routes a little more obvious 846
* And we've cleaned up test around some deprecated methods 847
* Don't forget that [gemd-python](https://github.com/CitrineInformatics/gemd-python/releases) also fixed some bugs with parsing complex unit expressions and serialization.

**Full Changelog**: https://github.com/CitrineInformatics/citrine-python/compare/v2.12.0...v2.15.0

In this release of Citrine Python, we are happy to announce some new changes to make developer's lives easier. We've included new predictor endpoints that allows folks to programmatically fix stale models, simplified some of our code around path construction for `file_link`s, and updated our allowed fingerprint methods in Generative Design. As always, we hope these changes keep you running smoothly.

What's New
* New endpoints to allow users to identify stale predictors (`predictors.is_stale(uid, version=version)`) and retrain stale predictors in-place (`predictors.retrain_stale(uid, version=version)`). This allows users to retrain predictors that have become stale and use single prediction features and Candidate Modification. 843

Improvements
* Simplified our path construction & escaping by centralizing and refactoring methods. 840
* We've updated our documentation and valid fingerprint types accepted in Generative Design strategies. 842

**Full Changelog**: https://github.com/CitrineInformatics/citrine-python/compare/v2.11.2...v2.12.0