Release 15.0 (2018-04-06)


This will be the last major release that supports Python 2. From release v16.0 on, only Python 3 will be explicitly supported. Minor (bugfix) releases of v15 will continue to support Python 2.

Source Identifier
Git tag 15.0
EUPS distrib v15_0

This release is based on the w_2018_10 weekly build.

These release notes highlight significant changes to the Science Pipelines codebase which are likely to be of wide interest. For a complete list of changes made, see Tickets Addressed in Spring 2018 Release (v15_0).

If you have questions or comments about this release, visit our Community Forum for advice.

See also:

Major Functionality and Interface Changes

Image differencing algorithm improvements

Algorithms appropriate for differencing in a variety of contexts have been implemented as tasks in the LSST stack. These various algorithms and configurations have been compared. The results are written up in DMTN-061. This represents a significant step forward in being able to determine the baseline image differencing algorithm. Relevant code changes are in the ip_diffim package.

Performance improvements in coaddition

Now all coaddition algorithms have significantly reduced memory footprints.

Significantly improved artifact rejection in coaddition

Coaddition algorithms that do artifact clipping can now handle artifacts that overlap from epoch to epoch. SafeClipAssembleCoaddTask and CompareWarpAssembleCoaddTask are the two examples.

Full focal plane sky subtraction

There are now tasks to create and apply models of the sky that extend over the entire field of view. View notes in the LSST Community form post.

Replace all Wcs classes with the AST backed SkyWcs

The last release introduced a transform system backed by the AST package. Since that release the stack has been converted to using that system in all contexts where a world coordinate system is required.

Plotting frontend for Firefly

This release includes, for the first time, the package that allows the LSST plotting abstraction layer to plot directly in the Science Portal plotting tool, Firefly. View display_firefly on GitHub.

Lossless compression on by default

Lossless compression is turned on by default when persisting any image-like data product. Read the decision and view the notes in this LSST Community forum post and links therein.

Changes to command-line tasks

Command-line tasks now handle clobbering of versions, data, and configs in a more intuitive way. E.g. output repositories are now expected to differ from the input repository. This eliminates the need to explicitly turn on clobbering when making multiple runs to different outputs (reruns) when using the same inputs. Additional details are in an LSST Community forum post.

Updated pre-requisites

Pre-requisites for installing the science pipelines have been updated. Of note are that numpy 1.13, astropy 2.0 and matplotlib 2.0 are all required. The baseline version of Python is now Python 3.6. See the announcement in this post on the LSST Community forum.