v2.1.0

Feature

• Lazy execution mode (#110) (314901e)
• Add EEOF inverse_transform() (#111) (ab6e3cc)

Documentation

• Remove dead files (6c4842e)

v2.0.3

Fix

• Remove poetry io group (#113) (23cca05)

v2.0.2

Fix

• Safe default zarr write mode (#107) (3003f5a)
• Safe default zarr write mode (0f95d04)

v2.0.1

Fix

• Circumvent pypi block with v2.0.1 release (#105) (dbad76a)
• Circumvent pypi block with v2.0.1 release (d014c87)

v2.0.0

Feature

• Serialization methods (#103) (69df4d9)
• Inverse transform for unseen data (#102) (4974844)
• Inverse_transform scores arg (7c1dc16)

Breaking

• inverse transform for unseen data (#102) (4974844)
• inverse_transform scores arg (7c1dc16)

Documentation

• Update authors and description (#100) (49c4367)
• Update authors and description (3e39eda)

v1.2.2

Fix

• Reindexing transformed scores (a8e458e)

Documentation

• Add performance tests (42d8e75)

Performance

• Decomposer: Compute by default (1f3b214)
• Dask: Avoid double computation (f24cb09)

v1.2.1

Fix

• Parameter to show dask progress bar (3482fa4)
• Bootstrapper: Correct scaling of PCs (ed59dd9)

Documentation

• Fix link (c6a595b)
• Minor tweaks in documentation (977a89e)
• Build with Python 3.11 (c181972)
• Complete overhaul (47e13ea)
• Add CoC (70f10b2)
• Update doc environment (#94) (9592dbc)
• Update env (e8f72d5)
• Repair gallery (fa647d8)
• Update examples (#93) (9f3c0c1)
• Update some docstrings (2795c54)
• Fix reference hyperlinks (38cadb9)
• Fix up minor errors (ef3fe8a)
• Fix broken example (ecc114e)
• Fix broken example (b58d985)

v1.2.0

Feature

• Provide standard kwarg for random_state (cf8b641)
• Add Extended EOF Analysis (2db736c)
• Parameter normalized scores (f0dcb34)
• Add GWPCA support (58c2dbe)
• Add CCA support (79e82cc)
• Choose name of sample and feature dim (4d521f8)

Fix

• CCA: Add checks for edge cases (76c0b1b)
• Streamline model attribute types (98c5cef)
• GWPCA: Raise error in scores (43e9274)
• Bootstrapper: Avoid pertubating sample coords (ffd11fc)
• Score coordinates in T-mode EOF analysis (731eec2)
• PCA preprocessing before Hilbert transform (2bbcb69)

Documentation

• Provide top-level type hints (6be8923)

Performance

• dask: Compute SVD result immediately (807b7e8)

v1.1.0

Feature

• Add Optimal Persistence Analysis (OPA) (2482110)
• Add Canonical Correlation Analysis (782e624)
• MCA: Provide optional PCA preprocessing (100f40d)
• Choose the SVD solver (50c2386)

Documentation

• More instructive quickstart (#75) (58c2f09)
• Correct typo in README (3b88d5e)
• Add quickstart example and point to methods (f741199)
• Update location of Action workflow for badge (650a8c8)
• Update location of Action workflow for badge (878c5f5)
• Update badges (51113e4)
• Add Black badge to README (759b011)

v1.0.5

Fix

• ListStacker: Use DA Stacker to unstack scores (c6479cf)
• MCA covariance (fraction) (3e4558d)
• Remove **kwargs from BaseModel classes (555d0db)
• Inverse_transform enforces real output (5e97160)
• Add missing conj for PCA inverse transform (e9d6fe1)
• Transform of rotator classes (3368ed0)
• Weigh modes with norm before reconstructing (13f42e0)

Documentation

• Remove interactivity (6e99b8e)
• Add mamba to RTD config (33d040b)
• Use mamba to create doc env (610facf)
• Add a Project description to conf.py (0a5cfe3)
• Improve documentation (ea32e7e)
• Extend and streamline doc strings of methods (a70b9d0)
• bootstrapper: Update to new design pattern (d61525b)
• Update changelog (ace7ac9)
• Add ipython to env to use syntax highlight (6b2d16b)

Performance

• Use xarray built-in method for variance (23af4d8)