v1.0.0
v1.0.0 Oct 12, 2021
- Enhancements
- Add support for creating EntitySets from Woodwork DataTables (#1277)
- Add
EntitySet.__deepcopy__
that retains Woodwork typing information (#1465) - Add
EntitySet.__getstate__
andEntitySet.__setstate__
to preserve typing when pickling (#1581) - Returned feature matrix has woodwork typing information (#1664)
- Fixes
- Fix
DFSTransformer
Documentation for Featuretools 1.0 (#1605) - Fix
calculate_feature_matrix
time type check andencode_features
for synthesis tests (#1580) - Revert reordering of categories in
Equal
andNotEqual
primitives (#1640) - Fix bug in
EntitySet.add_relationship
that causedforeign_key
tag to be lost (#1675) - Update DFS to not build features on last time index columns in dataframes (#1695)
- Fix
- Changes
- Remove
add_interesting_values
fromEntity
(#1269) - Move
set_secondary_time_index
method fromEntity
toEntitySet
(#1280) - Refactor Relationship creation process (#1370)
- Replaced
Entity.update_data
withEntitySet.update_dataframe
(#1398) - Move validation check for uniform time index to
EntitySet
(#1400) - Replace
Entity
objects inEntitySet
with Woodwork dataframes (#1405) - Refactor
EntitySet.plot
to work with Woodwork dataframes (#1468) - Move
last_time_index
to be a column on the DataFrame (#1456) - Update serialization/deserialization to work with Woodwork (#1452)
- Refactor
EntitySet.query_by_values
to work with Woodwork dataframes (#1467) - Replace
list_variable_types
withlist_logical_types
(#1477) - Allow deep EntitySet equality check (#1480)
- Update
EntitySet.concat
to work with Woodwork DataFrames (#1490) - Add function to list semantic tags (#1486)
- Initialize Woodwork on feature matrix in
remove_highly_correlated_features
if necessary (#1618) - Remove categorical-encoding as an add-on library (will be added back later) (#1632)
- Remove autonormalize as an add-on library (will be added back later) (#1636)
- Remove tsfresh, nlp_primitives, sklearn_transformer as an add-on library (will be added back later) (#1638)
- Update input and return types for
CumCount
primitive (#1651) - Standardize imports of Woodwork (#1526)
- Rename target entity to target dataframe (#1506)
- Replace
entity_from_dataframe
withadd_dataframe
(#1504) - Create features from Woodwork columns (#1582)
- Move default variable description logic to
generate_description
(#1403) - Update Woodwork to version 0.4.0 with
LogicalType.transform
and LogicalType instances (#1451) - Update Woodwork to version 0.4.1 with Ordinal order values and whitespace serialization fix (#1478)
- Use
ColumnSchema
for primitive input and return types (#1411) - Update features to use Woodwork and remove
Entity
andVariable
classes (#1501) - Re-add
make_index
functionality to EntitySet (#1507) - Use
ColumnSchema
in DFS primitive matching (#1523) - Updates from Featuretools v0.26.0 (#1539)
- Leverage Woodwork better in
add_interesting_values
(#1550) - Update
calculate_feature_matrix
to use Woodwork (#1533) - Update Woodwork to version 0.6.0 with changed categorical inference (#1597)
- Update
nlp-primitives
requirement for Featuretools 1.0 (#1609) - Remove remaining references to
Entity
andVariable
in code (#1612) - Update Woodwork to version 0.7.1 with changed initialization (#1648)
- Removes outdated workaround code related to a since-resolved pandas issue (#1677)
- Remove unused
_dataframes_equal
andcamel_to_snake
functions (#1683) - Update Woodwork to version 0.8.0 for improved performance (#1689)
- Remove redundant typecasting in
encode_features
(#1694) - Speed up
encode_features
if not inplace, some space cost (#1699) - Clean up comments and commented out code (#1701)
- Update Woodwork to version 0.8.1 for improved performance (#1702)
- Remove
- Documentation Changes
- Add a Woodwork Typing in Featuretools guide (#1589)
- Add a resource guide for transitioning to Featuretools 1.0 (#1627)
- Update
using_entitysets
page to use Woodwork (#1532) - Update FAQ page to use Woodwork integration (#1649)
- Update DFS page to be Jupyter notebook and use Woodwork integration (#1557)
- Update Feature Primitives page to be Jupyter notebook and use Woodwork integration (#1556)
- Update Handling Time page to be Jupyter notebook and use Woodwork integration (#1552)
- Update Advanced Custom Primitives page to be Jupyter notebook and use Woodwork integration (#1587)
- Update Deployment page to use Woodwork integration (#1588)
- Update Using Dask EntitySets page to be Jupyter notebook and use Woodwork integration (#1590)
- Update Specifying Primitive Options page to be Jupyter notebook and use Woodwork integration (#1593)
- Update API Reference to match Featuretools 1.0 API (#1600)
- Update Index page to be Jupyter notebook and use Woodwork integration (#1602)
- Update Feature Descriptions page to be Jupyter notebook and use Woodwork integration (#1603)
- Update Using Koalas EntitySets page to be Jupyter notebook and use Woodwork integration (#1604)
- Update Glossary to use Woodwork integration (#1608)
- Update Tuning DFS page to be Jupyter notebook and use Woodwork integration (#1610)
- Fix small formatting issues in Documentation (#1607)
- Remove Variables page and more references to variables (#1629)
- Update Feature Selection page to use Woodwork integration (#1618)
- Update Improving Performance page to be Jupyter notebook and use Woodwork integration (#1591)
- Fix typos in transition guide (#1672)
- Update installation instructions for 1.0.0rc1 announcement in docs (#1707
, :pr:
1708, :pr:
1713, :pr:
1716) - Fixed broken link for Demo notebook in README.md (#1728)
- Update
contributing.md
to improve instructions for external contributors (#1723) - Manually revert changes made by :pr:
1677
and :pr:1679
. The related bug in pandas still exists. (#1731)
- Testing Changes
- Remove entity tests (#1521)
- Fix broken
EntitySet
tests (#1548) - Fix broken primitive tests (#1568)
- Added Jupyter notebook cleaner to the linters (#1719)
- Update reviewers for minimum and latest dependency checkers (#1715)
- Full coverage for EntitySet.eq method (#1725)
- Add tests to verify all primitives can be initialized without parameter values (#1726)
Thanks to the following people for contributing to this release:
@bchen1116, @gsheni , @HenryRocha, @jeff-hernandez, @rwedge, @tamargrey, @thehomebrewnerd, @VaishnaviNandakumar