v0.3.3 (2025-11-21)

Fixes

• Updates prediction payload for pipeline service

v0.3.2 (2025-11-03)

Fixes

• Datetime parsing in new experiment validation.

v0.3.1 (2025-08-21)

Fixes

• Change new product code length validation (allow codes with length < 10)
• Fix dataset.get_data method for modified datasets

v0.3.0 (2025-05-14)

New Features

• Added support for different Process Formats (mammalian, microbial) in the SDK.

v0.2.6 (2025-05-13)

New Features

• Added support for startingIndex in prediction configuration. This allows specifying a custom starting index for predictions, enabling more flexible use cases.

Bug Fixes

• Fixed allowed types in prediction instances to correctly support string and bool for categorical variables.

Other Changes

• Minor improvements and adjustments in type definitions for better consistency and maintainability.

v0.2.5 (2025-01-28)

New Features

• Introduced a new method to retrieve full experiment data within a dataset.
• Added a new argument to the predict methods in Cultivation, enabling customization of Confidence Intervals for predictions.

Bug Fixes

• Fixed an issue where timestamp values were not correctly parsed to integers before data upload.

Other Changes

• Updated the prediction functionality to use a new endpoint and data structure while maintaining existing functionality.

v0.2.4 (2024-09-19)

Bug Fixes

• Add validations and formatting to variables of variant categorical and logical when importing new data.

v0.2.3 (2024-08-28)

New Features

• Add a new class FlowVariableReference to improve the creation of new Feeds / Flows Variables.

Bug Fixes

• Fixed data tabularization problem when uploading through the SDK.

Other Changes

• Add new validations.md file, a comprehensive breakdown of the validations performed by the SDK when creating a new entity or pushing new data.
• Improved the documentation on Flow Variable creation.

v0.2.2 (2024-07-24)

New Features

• Add verification options for SSL/TLS Certificates. The SDK now supports using custom CAs, self-signed certificates, and disabling SSL/TLS verification (docs).

Bug Fixes

• Fixed bug in formatting the Historical Model's prediction request body.

Other Changes

• Add Project URL to link PyPI to Github page