- Adding support for cell type queries
- Adding new functionality for getting max value of data type (and optionally gene/protein)
- Updating documentation and tests to include queries by modality
- Updating doc tests for greater independence from index version
- Updating doc tests, adding new query type
- Add new parameter to enable gene->dataset and protein->dataset queries
- Running tests on client test server
- Version 8 was skipped
- Support iteration over result sets.
- Add python environments to gitignore.
- Parameterize api endpoint doctests.
- Only use POST, for simplicity.
- Use values_included appropriately.
- Values in error are sorted.
- Run tests in parallel.
- Remove
get_list
from external interface. - Distinguish
ResultsSet
fromResultsList
. - Remove
values_type
parameter ResultsList
supports[]
- Generate pydocs, with accurate kwarg lists..
- Support queries for and by dataset
- Support queries for all entities by identifier
- Examples of query by cluster
- Example of query for cell by protein
- Reworking client-side parameter validation
- Integrating hubmap_cell_id_gen_py into examples
- Supporting queries for all entities of a particular type, including proteins
- Expanding table in readme
- Fewer preflight checks on requests: These are a maintenance burden.
- Remove complement magic method, and add difference.
- Require "has" param to be list, and update examples.
- Add subtypes of ResultsSet, and check that operands on set operations match.
- Issue templates
- Rename repo and package.
- Move to github CI
- Add ASCT-B to matrix in README
- Improve input validation logic
- Use slice syntax to get list
- Change to more fluent SDK:
select_TARGET(where='SOURCE', has='CRITERIA', ...)
- OO interface: Fewer details need to be repeated across methods, because they are kept in the object.
- Basic python interface: methods mirror API hits.