CHANGES.md

Release Notes

[Unreleased]

These significant changes are coming in light of the pyOpenSci review process. These changes primarily focus on the packaging and installation infrastructure. These enhancements do not impact the core codebase directly.

PetroFit Enhancements

• Transition to Hatch: Migrated from setuptools to hatch for streamlined packaging and dependency management. #209

• Configuration Enhancements #209:

  • Added .editorconfig to standardize code style across different editors.
  • Introduced GitHub issue templates in .github/ISSUE_TEMPLATE for consistent issue reporting.
  • Updated test workflow to utilize hatch for test environment setup.

• Docker Removal #209:

  • Removed Docker support, including deletion of the Dockerfile and the docker.py helper script.

• Legacy Files #209:

  • Removed legacy Astropy packaging infrastructure file _astropy_init.py.

• Project Configuration Updates #209:

  • Updated pyproject.toml to align with hatch configuration requirements.

• Deprecated requirements.txt, setup.cfg, setup.py, and tox in favor of new configuration management with hatch. #209

• Add badges #209:

  • Added a repostatus.org badge to the readme file.
  • A badge listing supported python versions.

• Add Generative AI Policy. #231

Version 0.5.1

August 16th 2024

PetroFit Enhancements

• Update installation requiements for use with python 3.12. #209

General bug fixes and small changes

• Update rtd to use ubuntu-22.04. #200
• update_default_config removed since astropy no longer needs it. #205
• Update rtd to Use Python 3.11.8. #206
• Update installation requiements for use with python 3.12. #209
• Remove astropy_helpers from requirements. #209

Version 0.5.0

August 28th 2023

PetroFit Enhancements

• Allow users to choose fitters in fit_model. #187
• Introduce TRFLSQFitter as the default fitter. #187
• Enable calc_uncertainties for fitting. This allows users to get paramter errors on fits. #187
• Oversampled PSF in PSFConvolvedModel2D. #187
• Remove CircleCI and add GitHub CI workflow. #181
• A new option to address low signal-to-noise ratio issues when measuring r_99 by allowing the use of r_80. While r_80 provides good estimates of r_50 (r_e), it can sometimes result in an overestimation of r_99. To ensure accuracy, correction grids can be utilized to estimate r_99 from r_80. However, we maintain the Petrosian measurement philosophy that correction grids should be reserved for high-precision measurements.
• Error estimates for all Petrosian Radii according to Strauss et. al 2002 Appendix A.3. #169
• Saving/reading profiles to/from file. #169
• Grid generation now has multi-processing support. #169
• Moffat2D and EllipMoffat2D Deprecated. #157
• PSFModel has been depreciated and replaced by PSFConvolvedModel2D. #156

General bug fixes and small changes

• Remove DecisionTreeRegressor. #192
• Change sklearn to scikit-learn. #174 by @astro-nova
• Remove kernel key word from segmentation functions since PhotUtils no longer accepts them. #169
• order_cat moved to segmentation sub-module. #169
• source_photometry moved to segmentation sub-module. #169
• make_radius_list moved to utils sub-module. #169
• Refactor: photometry.flux_to_abmag -> utils.hst_flux_to_abmag. #169
• Refactor: Move measure_fwhm from utils to modeling.fitting. #169
• Depreciate calculate_photometic_density. #169
• Depreciate object_photometry, now its called radial_photometry. #169
• Depreciate measure_fwhm. #169
• Change the bkg_sub keywords to bg_sub to match PhotUtils. #158
• Recenter example PSF #159

Version 0.4.1

August 22nd 2022

PetroFit Enhancements

• CITATION.rst added to provide citing and credit information. #130
• segm_mask can now take a list of labels which means it is possible to get multiple unmasked sources using masked_segm_image #143
• Limits of param bounds in get_default_sersic_bounds changed. #143
  • r_eff min bound set to 1e-3.
  • n min bound set to 0.1.
  • ellip max bound set to 0.99.
• Add GenSersic2D model, which is a Sersic model with generalized ellipse isophot (boxy) parameter. #145
• Implement PetroApprox which is a class with no-PSF polynomials which relate profile paramters such as C2080 and Sersic n. #146

General bug fixes and small changes

• Bug Fix: Example RMS image in docs was not the same size as the example data image. #114
• Bug Fix: Model attribute bug fix in PSFConvolvedModel2D. #145
• Print out PSF sum if it is not equal to 1. #145
• Bug Fix: Change make_catalog kernel_size parameter to odd number. #148

Version 0.4.0

March 26th 2022

Upgrades to fitting sub-module, see below for details.

PetroFit Enhancements

• Removed outdated petrofit.utils.cutout function. #93
• model_to_image function upgraded to use astropy.convolution.utils.discretize_model. #95
• PSFModel has been upgraded to PSFConvolvedModel2D.PSFConvolvedModel2D uses __init__ function to wrap models. PSFModel class has been deprecated and will be removed in v0.5. #100
• An origin parameter has been added to petrofit.models.make_grid. #100
• fitting.py and models.py have been moved into a new modeling folder. This allows for all modeling functions and classes to be imported as from petrofit.modeling import .... #100
• petrofit.modeling.fitting.fit_model can now accept fitting weights, so users can pass fitting weights using rms or error images. #100
• petrofit.modeling.fitting.plot_fit now displays 3 panel image of fits (input image, model-image, and residual). #100

Version 0.3.1

March 8th 2022

PetroFit Enhancements

• Removed kcorrect from Docker file. #77
• Added badges to README.rst.
• Implemented get_amplitude_at_r function that calculates the amplitude of an image at an isophotal radius. #81
• Removed outdated fit_plane function. #84
• Moved requirements from environment.yml to requirements.txt. Though both can be used to install PetroFit, environment.yml is recommended at this time. #72
• Upgraded circleci resource_class to medium+. #62

General bug fixes and small changes

• Cleaned up the docs, introduced a CHANGES.mdfile. #77
• Add release procedure for developers. #70