Skip to content

Releases: gplepage/gvar

gvar version 11.5.1

09 May 20:05
Compare
Choose a tag to compare

More tuning of gvar.evalcov, for better performance on dense matrices of
size 15x15 or larger. Also improved doc-strings for Cython-generated
functions.

gvar version 11.5

07 May 01:19
Compare
Choose a tag to compare

This update provides significant performance upgrades for gvar.evalcov (thanks again to Giacomo Petrillo) and gvar.evalcov_blocks.

gvar version 11.4

29 Apr 17:39
Compare
Choose a tag to compare

Significant rewrite of gvar.evalcov_blocks(g), making it much faster for large, dense covariance matrices. (Thanks to Giacomo Petrillo for important ideas.) Also gvar.gvar(x,cov) has new keywords, verify and fast, for checking the validity of and optimizing the covariance matrix cov.

gvar version 11.2

18 Mar 14:25
Compare
Choose a tag to compare

Minor fixes and updates (to gvar.ode and gvar.abs).

gvar version 11.1

01 Mar 20:29
Compare
Choose a tag to compare

Further updates to gvar.dump and gvar.load, giving increased flexibility.

gvar version 11.0

27 Feb 20:17
Compare
Choose a tag to compare

Much improved gvar.dump and gvar.load for serializing dictionaries, arrays, etc. containing GVar objects. Objects that contain other types of data, in addition to GVar objects, are now allowed, and the data can be stored in nested dictionaries, lists, arrays, tuples, etc. These new functions are meant to operate similarly to the corresponding functions in Python's pickle module.

gvar version 10.1.2

24 Feb 14:39
Compare
Choose a tag to compare

Fixes bug in gvar.dump introduced in version 10.1.

gvar version 10.1.1

23 Feb 18:54
Compare
Choose a tag to compare

Fixes inefficiency/bug in gvar.dump.

gvar version 10.1

23 Feb 05:34
Compare
Choose a tag to compare

Much improved gvar.dump and gvar.dumps. New functionality for gvar.deriv and gvar.GVar.deriv. New functions: gvar.is_primary, gvar.qqplot, and gvar.dependencies.

gvar version 10.0.3

07 Feb 18:40
Compare
Choose a tag to compare

Fixes issue with str(GVar) that was dropping significant digits in the mean when the standard deviation was very small. This only affected the string representation of a GVar, not its actual value.