Important breaking change:

Starting with v0.5.4, careless.cchalf will use intensities rather than structure factors by default. For users seeking to replicate the previous behavior, use the new --use-structure-factors flag when calling careless.cchalf.

In other programs, CC1/2 is typically calculated using intensities. In the past, careless.cchalf used structure factors, because they are what is estimated directly in the merging algorithm. This release changes the behavior of careless.cchalf to use the intensity column by default and be more consistent with other software.

What's Changed

• use intensities cc1/2. support abismal colnames by @kmdalton in #191

Full Changelog: v0.5.3...v0.5.4

This release includes a new feature enabling the double-Wilson prior r-value to be optimized during model training. This allows the user to get a fast estimate of the best parameter value. It also reverts the behavior of the scale bijector to a previous version. If users want to keep the same behavior as version 0.4.2 to 0.5.2, please use --scale-bijector=softplus rather than the default which is now --scale-bijector=exp.

What's Changed

• Fix dw by @kmdalton in #189
• Update README.md by @DHekstra in #186

New Contributors

• @DHekstra made their first contribution in #186

Full Changelog: v0.5.2...v0.5.3

Thanks to @hkwang, we have some improvements to the stats submodule. Thanks!

What's Changed

• reformat problematic plotting by @hkwang in #187

New Contributors

• @hkwang made their first contribution in #187

Full Changelog: v0.5.1...v0.5.2

This version adds clearer error messages when careless can't figure out what columns to use for metadata, intensities, uncertainties, and image identifiers.

What's Changed

• Add error messages for data keys by @kmdalton in #185

Full Changelog: v0.5.0...v0.5.1

What's Changed

• Removed reset_index() call causing DataFrame indices to be plotted by @PrinceWalnut in #178
• make cchalf use unweighted pearson by @kmdalton in #181

New Contributors

• @PrinceWalnut made their first contribution in #178

Full Changelog: v0.4.9...v0.5.0

This release simplifies install using TensorFlow probability 0.25 and TensorFlow version 2.18.0. The cuda install script should no longer be needed, but can still be used.

What's Changed

• set tf max version to 2.17 by @kmdalton in #176
• update pre-install script to tfp 0.25 by @kmdalton in #177

Full Changelog: v0.4.8...v0.4.9

This PR addresses a number of longstanding issues. The user facing changes are:

• Improved numerical stability for the --scale-bijector=exp setting
• Allow setting the y-axis limits for many of the careless stats scripts
• Enforce a consistent dmin between ASUs (#71 )
• Add a script for calculating I / SigI (#148)
• Make it clear that careless poly does not support .stream files

On the developer side:

• Adds a missing test for the Double Wilson prior

This release changes the default bijector for the standard deviation of the scale distribution from Exp to Softplus to increase numerical stability.

This release adds 3 new flags controlling gradient clipping during training.

• --clipvalue
• --clipnorm
• --global-clipnorm
  These each take floating point values and are disabled by default. The meaning of each is documented in the Keras Adam docs.

This version updates packaging and distribution without touching any user-facing functionality.