Version 1.4.6.3

Version 1.4.6.3 includes updates from Josh:

• tkinter is optional
• uses more random seeds so testing is more reproducible/reliable
• includes wheel files for Python 3.14

Thanks Josh!!

Version 1.4.6.1

This release includes a parallelization update for fdepsd. Parallelization can now be done via joblib by setting the 'parallel' parameter to "joblib". Many thanks to @cluffspace for this update!!

Version 1.4.5

No code changes for this version, but some nice updates for maintainability and usability:

• @joshayers added workflows to build wheel files automatically
• @morbult added a tweak for numba support that will help others avoid some conflicts

Many thanks!!

Version 1.4.4

Update to the build configuration. Huge thanks to @joshayers for this excellent update! :-)

* Reorganized package config to use pyproject.toml instead of setup.py.

* Added optional dependency groups and moved some linting config into pyproject.toml.

* Added unit-tests job.

* Updated CI job to run install in verbose mode.

* Use editable install for tests.

* Updated python-publish.yml job.

Version 1.4.3.8

Updated the pyyeti.nastran.bulk.asm2uset routine to accept try_rdextrn="pv_only".

• This option retains the full uset table and returns an additional output indicating which DOF are part of the a-set
• Many thanks to @djagodzinski!

Made very minor miscellaneous tweaks to the documentation.

Version 1.4.3.7

This is small update to fix an issue with NaNs in pyyeti.cyclecount.findap and pyyeti.locate.find_unique. Many thanks to @morbult for pointing this out! :)

Version 1.4.3.6

Added the "encoding" option to the read routines in pyyeti.nastran.bulk.

Version 1.4.3.5

Small update to fix a couple issues in datacursor: all-NaN vectors and ax[vh]line handling

Version 1.4.3.4

Add the "forms" option to pyyeti.nastran.bulk.wtextseout. The new default is now to mark the mass, stiffness, and damping matrices as symmetric.

Version 1.4.3.3

Release to fix encoding issue on Windows:

• update guitools.read_text_file to use UTF-8 encoding
• primarily impacts nastran.bulk.rd* routines

Many thanks to A. Amar for pointing out the encoding issue! :)