Changelog for Frames-map-reduce-0.4.1.1
v 0.4.1.1
- Bumped some upper bounds
v 0.4.0.0
- Added
splitOnDatafor assigning via a given set of data cols. - Bumped upper bounds
- Checked compilation w/ghc-8.10.2
- Fixed a redundant include warning for Monoid using CPP in Frames.Folds
v 0.3.0.0
- Added Combinator for aggregation in Frames.Aggregation along with helpers to create folds over data cols, promote simple functions to record functions to be used in aggregations, combine key aggregations.
- Added
toRecordFoldto the Folds modules to simplify making record to record folds from record to Type folds. See examples. - Added an example of its use in the example.
- Added the full record/functor generalization in Frames.Aggregation.General
- Added the specialization to
Maybein Frames.Aggregation.Maybe - Cleanup of unneccesary imports
- Suppressed Orphan Instance warnings about Hashable instances for
Recordand general record types. These instances should properly be in Vinyl. - Bumped some upper bounds for GHC 8.8. Still can't compile with 8.8+ until discrimination is updated.
v 0.2.0.0
- Added Combinators for
record (Maybe :. ElField) rs(Much thanks to Tim Pierson, @o1lo01ol1o, for the idea and the work!). - Added Combinators polymorphic in record type (
Rec,ARecorSRecare supported) and composed interpretation functorf :. ElField Maybe :. ElFieldcase now uses the general case in order to avoid code duplicationRecordcase still uses separate code because the amount of new coercion code required was enough to not seem worth the re-use of the polymorphic case.
v 0.1.0.2
- Changed example from "executable" to "test-suite". Now it doens't need to be built and its dependencies don't leak to library users.
v 0.1.0.1
- lowered containers lower bound
- rasied some outdated upper bounds (profunctors, hashable)
v 0.1.0.0
- Initial hackage release