Skip to content

Releases: bbfrederick/capcalc

Version 1.4.0 - 11/18/23 Container update

18 Nov 19:08
Compare
Choose a tag to compare
  • (Docker) Made compatible with the new (small) basecontainer architecture.

Version 1.3.9 - 4/25/23 omnibus checkpoint release

25 Apr 21:37
Compare
Choose a tag to compare
  • (package) Removed numba as a dependency.
  • (package) Made pyfftw an optional dependency (it's no longer in basecontainer).
  • (package) Made some proactive fixes for numpy 2.0.0 compatibility.
  • (package) Accepted several dependabot PRs.
  • (package) General cleanup of automated workflows to match updates to rapidtide.
  • (Docker) Cleaned up and simplified Dockerfile.
  • (Docker) Updated to basecontainer:latest-release (0.3.5 atm).

Version 1.3.7 - 9/13/23 checkpoint release

13 Sep 14:10
Compare
Choose a tag to compare
  • Mass merge of a bunch of dependabot PRs.
  • (Docker) Updated to basecontainer 0.2.3.

Version 1.3.6 - 5/11/23 updated to python 3.11

12 May 02:29
Compare
Choose a tag to compare
  • (Docker) Updated to python 3.11 basecontainer.
  • (Docker) Fixed testdocker.sh.
  • (package) Modernized install procedure.

Version 1.3.5 - 1/11/23 checkpoint

11 Jan 20:07
Compare
Choose a tag to compare
  • (capfromtcs) Fixed summary statistics outputs.

Version 1.3.4 - 1/10/23 emergency bug fix

11 Jan 00:19
Compare
Choose a tag to compare
  • (capfromtcs) Revert change to max_iter when specifying initialcenters.

Version 1.3.3 - 1/10/23 checkpoint release

10 Jan 20:59
Compare
Choose a tag to compare
  • (capfromtcs) Changed intialization of kmeans from initialcenters, save more intermediate results.
  • (utils.py) Factored out some transition array calculations.

Version 1.3.2 - 11/15/22 emergency patch

15 Nov 14:45
Compare
Choose a tag to compare
  • (capfromtcs) Fixed a typo in specifying the initial k-means++ method.
  • (io) Updated libraries to match some changes in rapidtide.

Version 1.3.1 - 11/10/22 Alternative method of reusing old models

10 Nov 23:35
Compare
Choose a tag to compare
  • (capfromtcs) Added new option "--initialcenters" to allow specification of k-means cluster centers. This lets you reuse the CAPS as previous runs, but gets around the difficulty of reading back old models run under a different version of python or scikit-learn.

Version 1.3.0 - 11/10/22 Major code cleanup and bug fix

10 Nov 20:40
Compare
Choose a tag to compare
  • (package) Removed rapidtide as a dependency by copying over the necessary support routines.
  • (Docker) Updated basecontainer to latest, switched over to using pip rather than mamba for env.
  • (docs) Corrected a problem that was causing readthedocs build to fail.