Changelog¶
deeporigin v3.0.0
¶
Breaking changes
This release is a breaking change. Responses to all functions are no longer objects, but are now dictionaries wrapped in python Boxes.
Bugfixes¶
- Fixed a bug where writing more than 1000 rows failed
- Fixed a bug where constructing a dataframe from an empty DB failed
New features¶
- Ability to add new data to existing dataframes
- Support for Deep Origin platform SDK
- Deep Origin dataframes now disallow adding rows
- Better support for writing data back to databases
Miscellaneous¶
- Initial support for integration tests with Deep Origin Data Hub
- Large refactor of testing
deeporigin v2.4.0
¶
Bugfixes¶
- fixed a bug in how tokens were refreshed
- fixed a bug in CLI config use
New features¶
- added the ability to download files in parallel
- added ability to upload files with file hashes
- added the ability to decode user tokens
- support for a async client for the Deep Origin Data Hub API
- reference IDs now tracked in dataframes
- dataframes now support expressions and users
- dataframe slicing now restricts addition of new rows to slice
- dataframe updates more atomic: a single cell change only triggers updating that cell
- ability to delete terminated workstations
- package wide ability to check for latest versions on PyPI
Miscellaneous¶
- testing now includes python 3.13
- Data Hub API route now set to
datahub-api/api/
- improved test coverage
deeporigin v2.2.1
¶
- added the ability to interact with the platform, including
- fetching user secrets from the platform
- fetching user info from the platform
- fetching workstation info from the platform
- removed support for graphQL endpoints for secrets and variables
- support for new (breaking) syntax for deleting databases, folders, rows, and columns
- better support for caching tokens in filesystem
- numerous bug fixes and small improvements
- dropped support for Python 3.9
- better syntax for interacting with the
api.list_files
function - Improvements to Deep Origin DataFrames, including:
- ability to create new columns in dataframes and push them to a database
- DataFrames now show names of user who created and last updated a database, not just a ID
- DataFrame pretty printing methods now never interrupt core logic, and fall back to printing methods of superclass on failure
- DataFrames now support
df.tail()
anddf.head()
deeporigin v2.1.1
¶
- fixed a bug where dataframe header links were sometimes broken
- improved display of dataframe headers
- improved typing in dataframes
- added ability to check for newest versions on PyPI
- version numbers are now PEP 440 compliant
deeporigin v2.1.0
¶
- dropped support for column keys
- added the deeporigin.DataFrame class, which:
- is a drop-in replacement for pandas.DataFrame
- supports all pandas methods
- supports automatic syncing with a Deep Origin database