LiberTEM 0.14.0 and LiberTEM-blobfinder 0.6.0/0.6.1 released

[Dieter Weber]

Dear subscribers,

I am happy to announce the release of LiberTEM version 0.14.0 and
LiberTEM-blobfinder version 0.6.0/0.6.1

The minimum supported Python version was increased to 3.9 in line with a
range of other projects. Older versions of LiberTEM remain available in
situations where Python cannot be upgraded.

The modules libertem.analysis.gridmatching and
libertem.analysis.fullmatch are now available from
libertem_blobfinder.common.gridmatching and
libertem_blobfinder.common.fullmatch. The original import locations
trigger a deprecation warning since 0.14.0 and are slated for removal in
LiberTEM 0.16.

## LiberTEM 0.14.0

* System for UDFs to be aware of which parts of a dataset have already
been processed during merging and post-processing, and to mark parts of
results as invalid when the processing would otherwise be incorrect:
https://libertem.github.io/LiberTEM/udf/advanced.html#valid-data-masking
* Fix the CoMUDF automatic offset and gradient compensation to yield
reasonable results during live processing. Thank you Sivert for the test
case!
* Fix loading of EMPAD data acquired in “search” mode. Thank you Cigdem
for providing test data!
* Executor snoozing, --snooze-timeout command line option: Stop workers
after a period of inactivity, useful for deploying LiberTEM as a service
on multi-user systems. See libertem-server --help or
https://libertem.github.io/LiberTEM/usage.html#configuring-the-libertem-server-advanced
for all command-line options.

See https://libertem.github.io/LiberTEM/changelog.html#v0-14-0 for the
full change log!

## LiberTEM-blobfinder 0.6.0/0.6.1

The no-change 0.6.1 release was created because the Zenodo upload failed
for 0.6.0.

* Fourier upsampling is now implemented for calculating the ‘refineds’
peak positions with the upsample parameter, see
https://libertem.github.io/LiberTEM-blobfinder/reference.html#libertem_blobfinder.udf.refinement.run_refine
* Correlation UDFs now support GPU processing with cupy as well as
sparse array input data.

See https://libertem.github.io/LiberTEM-blobfinder/changelog.html#v0-6-1
for the full changelog!

Many thanks to everybody who contributed to these releases!

## Future plans

* Work on https://github.com/LiberTEM/LiberTEM-live-server for better
live processing experience; blend live and offline processing. Special
thanks to Sivert and Magnus for working towards convenient live CoM with
descan error compensation!
* Properly release https://github.com/LiberTEM/Microscope-Calibration
* Extend descan error compensation to more UDFs; unify interfaces.
* Improved performance for descan error compensation.
* Dedicated UDFs for ring and disk virtual detectors similar to the
CoMUDF for easier live processing.
* Performance optimization for sparse data and GPU processing.

Please let us know what features are relevant for you so that we can
prioritize accordingly!

You can reach us online in the Issue tracker
<https://github.com/LiberTEM/LiberTEM/issues>, Gitter chat
<https://gitter.im/LiberTEM/Lobby> or by e-mail: Alexander Clausen
<a.cl...@fz-juelich.de>, Dieter Weber <d.w...@fz-juelich.de>, Matthew
Bryan <matthe...@cea.fr>.

With best regards,
Dieter


--

Dr. Dieter WEBER

Ernst Ruska-Centre for Microscopy and Spectroscopy with Electrons (ER-C)
Forschungszentrum Jülich
52425 Jülich, Germany

Email: d.w...@fz-juelich.de
Phone: +49 2461 61 85118