Skip to content

Latest commit

 

History

History
104 lines (81 loc) · 3.79 KB

CHANGES.rst

File metadata and controls

104 lines (81 loc) · 3.79 KB
Raw

v0.4 (February 26, 2024)

  • Upgrade Python dependencies.

v0.3.5 (August 02, 2022)

  • Use the new from_dataset method to initialize objects in examples. (#123)
  • Add HBN (#121)

v0.3.4 (August 02, 2022)

  • Add Python 3.10 to CI workflows (#122)
  • Add fit, fit_transform, transform, predict, and score methods (#119)
  • NF: Allow passing an AFQDataset object as input. (#113)
  • Add 1D cnn models and data augmentation (#112)
  • (#107)
  • Remove groupyr ref from user guide (#110)
  • Remove groupyr references from contribution guide (#109)

v0.3.3 (February 16, 2022)

  • Fix attribute errors in bagging and cross-validation. Loosen dependency reqs (#108)
  • Make AFQDataset work with different input formats and make indexable (#105)

v0.3.2 (December 22, 2021)

  • ENH: Expose figsize and tight_layout kwargs to plot_tract_profiles (#104)
  • Expose legend kwargs to user in plot_tract_profiles (#102)

v0.3.1 (December 07, 2021)

  • BF: Fix label encoding in the presence of NaN labels (#101)
  • Add enforce_sub_prefix option to AFQDataset and load_afq_data (#96)
  • RF: For plotting, sometimes you don't want to group by any other variable (#99)
  • BF: Cast integer session identifiers to strings. (#91)
  • Add an AFQDataset class (#89)
  • Add callosal bundles to plot_tract_profiles (#82)
  • ENH: Add Convolutional Neural Network Models (#83)
  • TST: Improve coverage. (#86)
  • Fixes tests broken by previous commits. (#85)
  • DOC + RF: Adds docstrings to class definitions. (#84)
  • ENH: Allow bundle aggregation and put load_afq_data output into a named_tuple (#80)
  • ENH: Allow bundle aggregation and put load_afq_data output into a named_tuple (#80)
  • DOC: Make examples a bit more explicit (#79)

v0.3.0 (June 28, 2021)

  • Add research examples to documentation (#77)
  • ENH: Add dataset fetchers (#76)
  • Adds error handling for cases where users forget to set target_cols. (#73)
  • ENH: Add unsupervised boolean parameter to load_afq_data (#71)
  • ENH: Add bundle profile plotting function (#69)
  • Remove transformers that were migrated to groupyr (#68)
  • DEP: Loosen groupyr dependency (#67)
  • DEP: Loosen dependency requirements (#66)
  • Fixes broken links (#65)
  • ENH: Stratify bootstrap replicants in SerialBaggingClassifier (#61)
  • ENH: Add SerialBagging classes (#60)

v0.2.8 (December 10, 2020)

  • DEP: Bump groupyr version

v0.2.7 (December 09, 2020)

  • ENH: Add ensemble estimators to pipelines (#59)

v0.2.6 (October 28, 2020)

  • FIX: Add option to serialize cv execution (#57)

v0.2.5 (October 26, 2020)

  • ENH: Use joblib.Parallel instead of custom groupyr._ProgressParallel (#56)

v0.2.4 (October 26, 2020)

  • ENH: Add functions and test for outer cross-validation loop (#55)

v0.2.3 (October 22, 2020)

  • FIX: Use target transformer only if requested by user (#54)

v0.2.2 (October 15, 2020)

  • ENH: Bump groupyr version (#53)
  • ENH: Use sklearn LabelEncoder in load_afq_data and return classes (#52)
  • ENH: Increase flexibility of load_afq_data (#51)

v0.2.1 (October 14, 2020)

  • Major refactoring after moving the sparse group lasso estimator functions into a new library (groupyr) and retaining only AFQ specific dataset transformers and pipelines in AFQ-Insight

v0.2.0 (October 14, 2020)

  • Major refactoring after moving the sparse group lasso estimator functions into a new library (groupyr) and retaining only AFQ specific dataset transformers and pipelines in AFQ-Insight