Autofit

- Starting point API for starting an MCMC fit with walkers in certain positions or maximum likelihood estimator fit with a start point implemented (https://github.com/rhayes777/PyAutoFit/pull/562). The example tutorial script for this feature is not written yet.

- Dynamic delta scaling in expectation propagation fits, which prevent over confident result due to error underestimation (https://github.com/rhayes777/PyAutoFit/pull/559). The example tutorial script for this feature is not written yet.

- Faster generation of models by sampling better within the prior limits (https://github.com/rhayes777/PyAutoFit/pull/558).

- Can make a parameter free across al combined analysis objects (docs / cookbook to be written):


analysis = sum(analysis_list)
analysis = analysis.with_free_parameters(
model.parameter,
)


- Model composition using relations (cookbook to be written):


x_list = [464, 658, 806]

m = af.UniformPrior(lower_limit=-0.1, upper_limit=0.1)
c = af.UniformPrior(lower_limit=-10.0, upper_limit=10.0)

analysis_list = []

for x, imaging in zip(x_list, imaging_list):

y = af.Add(af.Multiply(x, m), c)

analysis_list.append(
al.AnalysisImaging(dataset=imaging).with_model(
model.replacing(
{
model.gaussian.x: gaussian.x,
}
)
)
)


- Tutorials for fitting a hierarchical model outside of EP.

https://github.com/Jammy2211/autofit_workspace/tree/release/notebooks/howtofit/chapter_graphical_models

- Stability upgrades to EP framework.

- Support for Python 3.9, 3.10.
- `LogGaussianPrior` implemented.
- Simultaneous fitting of hieraerchical models (E.g. not just via EP) supported.
- Minor updates to graphical model API.

- Sensitivity mapping now have options for customizing priors on the sensitivity component.
- New API for aspects of graphical models.

This release primarily includes a lot of continued develop of the graphical modeling framework:

https://pyautofit.readthedocs.io/en/latest/features/graphical.html

There are now 4 fully functional tutorials on graphical models in the autofit_workspace, which include expectation propagation and hierarchical models:

https://github.com/Jammy2211/autofit_workspace/tree/release/notebooks/howtofit/chapter_graphical_models

The release contains a lot of small improvements and additional features to the database, search grid search and general model-fitting. These are documented throughout the autofit_workspace.

We have switched to a date-based numbering system, with a long-term to do an overnight build nightly.

- Add scipy LBFGS as optimizer.
- Fixed bugs where building of database via scrape method would not support all queries.
- Database and Aggregator support for GridSearch objects and results.
- Improvements to unique identifier including option to exclude certain values from being tracked.
- Interal storage of mappings between priors and model now uses the string representation of the prior with an id, to ensure there is no ambuigity in matching,
- Further development on graphical modeling framework, primarily refactoring of existing code.