Erlab

🐞 Bug Fixes

- **io:** make summary caching togglable ([99b8e22](https://github.com/kmnhan/erlabpy/commit/99b8e221e75db73382bf599170c58d8a68ca049e))

Also fixes a bug where interactive summary plots were duplicated

- **io:** data loader related fixes ([da08e90](https://github.com/kmnhan/erlabpy/commit/da08e9076e59895b35c393c8e2556c3592adf4a5))

DA30 dataloader now preserves case for attribute names from zip files.

Post processing for datasets now works properly

🐞 Bug Fixes

- **interactive:** keep pointer for imagetool, fix typing issues ([c98c38e](https://github.com/kmnhan/erlabpy/commit/c98c38ea11bce50ed9bfd8d374064bb2b1659d0c))

♻️ Code Refactor

- move `characterization` to `io` ([9c30f1b](https://github.com/kmnhan/erlabpy/commit/9c30f1b7df51460f502dcbf999e3fac34be1cf99))

✨ Features

- **kspace:** rewrite conversion with `xarray.apply_ufunc` ([156cef8](https://github.com/kmnhan/erlabpy/commit/156cef830582e01dc378a7437a0c85f4c7efc077))

Momentum conversion now relies on xarray broadcasting for all computations, and objects with extra dimensions such as temperature can be automatically broadcasted.

Dask arrays can also be converted.

- **exampledata:** enable specifying seed for noise rng ([aa542e8](https://github.com/kmnhan/erlabpy/commit/aa542e8c288ff1ca64820960f469b2c244ca5c95))

- **interpolate:** enable fast interpolation for 1D arrays ([ff333a0](https://github.com/kmnhan/erlabpy/commit/ff333a05803d7079034e36f2e1dc3d22d0b686f7))

- make both arguments optional for loader_context ([6780197](https://github.com/kmnhan/erlabpy/commit/6780197f68abfe7a9edbda951d804a9bc5ba56e9))

- **kspace:** automatically detect kinetic energy axis and convert to binding ([bbde447](https://github.com/kmnhan/erlabpy/commit/bbde44717155d1dd9ffefbc286da32b0bfac2180))

- add more output and parallelization to fit accessor ([59b35f5](https://github.com/kmnhan/erlabpy/commit/59b35f53f3ef7f518aec92e05854dba42ddba56f))

Allows dictionary of `DataArray`s as parameter to fit accessor.

Now, the return `Dataset` contains the data and the best fit array. Relevant tests have been added.

- add callable fit accessor using apply_ufunc ([11e3546](https://github.com/kmnhan/erlabpy/commit/11e35466fec158e40d0e8e738dd81ed10225d83c))

Add a `Dataset.modelfit` and `DataArray.modelfit` accessor with similar syntax and output as `Dataset.curvefit`. Closes [22](https://github.com/kmnhan/erlabpy/issues/22)

- add option to plot_array_2d so that users can pass non-normalized color array ([74cf961](https://github.com/kmnhan/erlabpy/commit/74cf961532a50d9c324189318460a9f840291a85))

- **analysis.gold:** add option to normalize energy axis in fitting ([3dffad6](https://github.com/kmnhan/erlabpy/commit/3dffad65993520c4b9a9a3afd6be85671bac9d3a))

This improves performance and results when eV is large like ~100eV.

🐞 Bug Fixes

- **kspace:** allow explicit coordinate kwargs ([fe47efc](https://github.com/kmnhan/erlabpy/commit/fe47efcde941767c02b582ce8b29d4b3678fd843))

- **exampledata:** change noise generation parameters ([b213f11](https://github.com/kmnhan/erlabpy/commit/b213f1151ed2555fc80374e9ebe3fc0856a13948))

- **fit:** make FermiEdge2dModel compatible with flattened meshgrid-like input arrays ([c0dba26](https://github.com/kmnhan/erlabpy/commit/c0dba261670774862f2dfae62c770bbab81aac2f))

- fix progress bar for parallel objects that return generators ([23d41b3](https://github.com/kmnhan/erlabpy/commit/23d41b31a3f3ee6c7343d471f7cec34dc374bafa))

Tqdm imports are also simplified. We no longer handle `is_notebook` ourselves, but just import from `tqdm.auto`

- **plotting:** fix 2d colormaps ([8299576](https://github.com/kmnhan/erlabpy/commit/8299576ce3cbcbaec106bef952c6df148bb7ca18))

Allow images including nan to be plotted with gen_2d_colormap, also handle plot_array_2d colorbar aspect

♻️ Code Refactor

- make zip strict (ruff B905) ([78bf5f5](https://github.com/kmnhan/erlabpy/commit/78bf5f5a2db52c14ccf5bfd3c83659ca53c4a408))

- fix some type hints ([2dfa5e1](https://github.com/kmnhan/erlabpy/commit/2dfa5e1b4582e00d0631376ee32aa7d0b1b945b6))

- **example:** move exampledata from interactive to io ([1fc7e6c](https://github.com/kmnhan/erlabpy/commit/1fc7e6c22ce477fe7ebbd8b0c6844d1a85df3fcf))

Also add sample data generation for fermi edge

- refactor accessors as submodule ([9fc37bd](https://github.com/kmnhan/erlabpy/commit/9fc37bd4825de519e4c4b6e30e9e32bf9392ed2d))

- rewrite either_dict_or_kwargs with public api ([34953d1](https://github.com/kmnhan/erlabpy/commit/34953d10b6fd67720b1c29dbed1ab7a24e4d3060))

- move correct_with_edge from era.utilities to era.gold ([08a906f](https://github.com/kmnhan/erlabpy/commit/08a906ff61a74febc0f47ed08ac24e7a4cd0977f))

Calling from utilities will now raise a DeprecationWarning.

The erlab.analysis namespace is unchanged, so the affect will be minimal.

- qsel now raises a warning upon scalar indexing outside coordinate bounds ([d6ed628](https://github.com/kmnhan/erlabpy/commit/d6ed628111be8ac594d3a1b83cc2785a31e3f06e))

🐞 Bug Fixes

- **io:** unify call signature for summarize ([e2782c8](https://github.com/kmnhan/erlabpy/commit/e2782c898d5aaaa1443b2bc82bb61fb40a28d232))

- resolve failing tests due to changes in sample data generation ([80f0045](https://github.com/kmnhan/erlabpy/commit/80f004574950834e42dbfa7677031d0f9f113bda))

- **interactive.exampledata:** properly generate 2D data ([825260c](https://github.com/kmnhan/erlabpy/commit/825260c8ceb0a79b8c071750003529b91cda3573))

⚡️ Performance

- **io:** speedup merlin summary generation by excluding duplicates ([d6b4253](https://github.com/kmnhan/erlabpy/commit/d6b42537ce48232b5112daef8f31e5cf86ea921a))

♻️ Code Refactor

- **io:** allow for more complex setups ([f67b2e4](https://github.com/kmnhan/erlabpy/commit/f67b2e4c7b092b7ca2db00ce02a23647879c514b))

LoaderBase.infer_index now returns a second argument, which is a dictionary containing optional keyword arguments to load.

- **io:** provide rich interactive summary ([b075a9e](https://github.com/kmnhan/erlabpy/commit/b075a9ee59b61892462fc475e78b036a54408099))

- **io:** include "Path" column in ssrl loader summary ([ae1d8ae](https://github.com/kmnhan/erlabpy/commit/ae1d8aee051aa71563f6a6009ce9672e56edfae7))

- **io:** improve array formatting in summary ([1718529](https://github.com/kmnhan/erlabpy/commit/171852957db7fe53ff6a5c5c5f843530078d4b46))

🐞 Bug Fixes

- **fit:** add sigma and amplitude expressions to MultiPeakModel parameters ([3f6ba5e](https://github.com/kmnhan/erlabpy/commit/3f6ba5e84922129296183e02255506df73da0276))

- **fit.minuit:** properly handle parameters constrained with expressions ([d03f012](https://github.com/kmnhan/erlabpy/commit/d03f012b4fde92f445a24657dca1fb5b3600fa45))

♻️ Code Refactor

- set informative model name for MultiPeakModel ([d14ee9d](https://github.com/kmnhan/erlabpy/commit/d14ee9d6ac7962207700de50039a5b7a858fea6a))

- add gaussian and lorentzian for consistency ([07c0dfb](https://github.com/kmnhan/erlabpy/commit/07c0dfb9ecfb882e4f5f0ccfe942c1a835b613b2))

✨ Features

- enable component evaluation for MultiPeakModel ([8875b74](https://github.com/kmnhan/erlabpy/commit/8875b7443d26313156fcdcc43586d40af4ff4f00))

- **analysis.fit:** add BCS gap equation and Dynes formula ([f862aa4](https://github.com/kmnhan/erlabpy/commit/f862aa4af4d2ba470f1ea074fc90442d9b18b336))

🐞 Bug Fixes

- curvefittingtool errors ([9abb99c](https://github.com/kmnhan/erlabpy/commit/9abb99c35633bc722469276d4837a2372c132042))

♻️ Code Refactor

- cleanup fit namespace ([906aa99](https://github.com/kmnhan/erlabpy/commit/906aa99193f78577e705218b2d6c22378611f84b))

- rename ExtendedAffineBroadenedFD to FermiEdgeModel ([a98aa82](https://github.com/kmnhan/erlabpy/commit/a98aa82bcbdf22ff8a156d800e336653f9afba07))

- **interactive:** exclude bad colormaps ([877c915](https://github.com/kmnhan/erlabpy/commit/877c915def6eb3dddb3862d6ac64c8c70f456ad3))