Releases: CUQI-DTU/CUQIpy
Releases · CUQI-DTU/CUQIpy
v1.3.0
What's Changed
- Add gradient to uniform distribution by @chaozg in #526
- Fix typo (PCN intead of pCN in experimental) in BayesianProblem by @chaozg in #534
- update NUTS acc rate list update by @amal-ghamdi in #536
- fix NUTS acceptance rate by @amal-ghamdi in #539
- add test for counting accepted samples in NUTS by @amal-ghamdi in #540
- Checkpointing nuts by @amal-ghamdi in #530
- Display acc rate on progress bar by @chaozg in #533
- update links in README by @chaozg in #523
- Remove pre_sample, pre_warmup from NUTS and fix warning by @amal-ghamdi in #545
- Add demos for use of truncated Normal by @chaozg in #549
- Add (iid) truncated normal and also add grad to normal by @chaozg in #542
- Add ADMM + parameter reordering FISTA by @jeverink in #553
- creation of a MYULAnew class by @remilaumont in #407
- remove file added by mistake by @amal-ghamdi in #565
- Update repr method in Sampler class by @amal-ghamdi in #566
- Add AffineModel by @siljachr in #486
- fix copy, prior name in MYULA + docs artifacts by @amal-ghamdi in #567
- Add how-to on Image2D by @chaozg in #572
- Regularized Gaussian extension: TV by @jeverink in #564
- Add support for algebra by @nabriis in #583
- Make the interface of restorator more robust by @chaozg in #589
- Add random variable class by @nabriis in #597
- Refine existing solvers by @chaozg in #598
- add implementation of _ProductGeometry by @amal-ghamdi in #593
- Add support for conditioning on random variables by @nabriis in #600
- Use scipy.optimize.lsq_linear in RegularizedLinearRTO for nonnegativity by @chaozg in #602
- Update _geometry.py with correct link to howto on image2d by @nabriis in #590
- Regularized Gaussian: TV with Gibbs by @jeverink in #587
- Add sample method to RV by @nabriis in #603
- Regularized Gaussian: Combined regularized and constraint presets by @jeverink in #612
- Add rv method to Distribution class and update artifact version by @chaozg in #611
New Contributors
- @remilaumont made their first contribution in #407
- @siljachr made their first contribution in #486
Full Changelog: v1.2.0...v1.3.0
v1.2.0
What's Changed
- Ensure constants are stored in densities when reducing to single density in joint distribution by @nabriis in #482
- Replace progressbar2 by tqdm by @chaozg in #502
- Fix plot of 1D geometries by @chaozg in #497
- Add 479 experimental module in bayesian problem by @nabriis in #500
- move tqdm from requirement-dev to requirements by @chaozg in #510
- Improve (hopefully) how HybridGibbs uses samplers state and history by @nabriis in #503
- Update documentation by @chaozg in #508
- Update documentation on experimental mcmc module by @nabriis in #513
- Add prefix to sampling progress by @nabriis in #520
- Update example in HybridGibbs by @nabriis in #521
- Fix issue of invalid proposals being accepted by @chaozg in #518
Full Changelog: v1.1.1...v1.2.0
v1.1.1
What's Changed
- Remove postfix New (only) from sampler names by @chaozg in #477
- update installation instructions and use with vscode by @chaozg in #480
- Fix HybridGibbs using initial point from sampling strategy by @nabriis in #484
- Add repr method for cuqi samples by @chaozg in #472
- Sprint34 add 1D, 2D utility plot function by @amal-ghamdi in #491
Full Changelog: v1.1.0...v1.1.1
v1.1.0
What's Changed
- Sampler rework via MCMC module (experimental module) by @chaozg in #356
- Make progressbar2 dependency not explicit by @nabriis in #385
- Sampler rework ULANew by @chaozg in #380
- Sampler rework: refactor CWMH sampler by @amal-ghamdi in #376
- fix arviz import by @amal-ghamdi in #396
- Sampler rework LinearRTONew and RegularizedLinearRTONew by @chaozg in #382
- Implement get_state, set_state, get_history, set_history in base class. by @nabriis in #390
- Sampler rework: NUTS by @amal-ghamdi in #387
- TEST: Sampler rework ugla 2 by @amal-ghamdi in #411
- Sampler rework UGLA by @chaozg in #405
- reduce time on zexperimental tests by @chaozg in #416
- Let Gaussian accept LinearOperator represented sqrtprec by @chaozg in #420
- Add tune to pCN and rename pCN to PCN by @chaozg in #418
- Allow samplers to be initialized without target by @nabriis in #408
- Add Gibbs sampler by @nabriis in #409
- Add (Hybrid) Gibbs sampler by @nabriis in #425
- Regularized Gaussian extension: ModifiedHalfNormal by @jeverink in #429
- Find valid samplers by @jeverink in #446
- Regularized Gaussian extension: RegularizedUniform and Constraints/Nonnegative Gaussian/GMRF alternatives by @jeverink in #430
- Add SmoothedLaplace by @chaozg in #450
- General grad check in userdefined distribution by @amal-ghamdi in #457
- Improve Conjugacy safety and flexibility by adding more validation by @nabriis in #444
- Conjugacy safety and flexibility by @jeverink in #359
- correct math display error in smoothed Laplace by @chaozg in #459
Full Changelog: v1.0.0...v1.1.0
v1.0.0
What's Changed
- Allow lazy import of arviz by @nabriis in #327
- Refactor GMRF and remove physical dim parameter by @nabriis in #333
- Fix missing cookie png by @nabriis in #336
- update plot_ci to make it more general by @amal-ghamdi in #338
- Fix bug in distribution gallery by @chaozg in #340
- Trivial fix to improve FD gradient computation performance by @amal-ghamdi in #353
- Implemented Implicited Regularized Gaussians by @jeverink in #291
- Make Gamma distribution multivariate by @nabriis in #358
- Sprint27 beta grad by @amal-ghamdi in #363
- Distributions: Disable conditioning on mutable variables if not conditioning variable by @nabriis in #361
New Contributors
Full Changelog: v0.8.0...v1.0.0
v0.8.0
v0.7.0
What's Changed
- change developers to maintainers by @chaozg in #288
- change get_components to regular method by @chaozg in #302
- Bump minimum supported python version, add latest python versions to testing and update maintainers in toml by @nabriis in #306
- fix get_indirect_variables as suggested in issue #289 by @chaozg in #309
- update demos/demo35_FD_gradient.py to test NUTS and ULA with FD approximated gradients by @chaozg in #307
Full Changelog: v0.6.0...v0.7.0
v0.6.0
What's Changed
- Sprint20 fix warnings by @amal-ghamdi in #270
- add ability to store run diagnostics for inspection (NUTS sampler) by @amal-ghamdi in #260
- Sprint20 fix geom multiplot by @amal-ghamdi in #271
- Add cookie phantom and update Deconvolution2D to use cuqi.data by @nabriis in #275
- rename Linear_RTO to LinearRTO by @amal-ghamdi in #284
- Fix 277 update readme and landing page by @nabriis in #285
- Add list of developers by @chaozg in #287
New Contributors
Full Changelog: v0.5.0...v0.6.0
v0.5.0
What's Changed
- Sprint19 time sol pde by @amal-ghamdi in #224
- Update failing unit tests by @nabriis in #252
- Sprint20 fix detect cond dist by @nabriis in #249
- Use geometry to define shape of distributions by @nabriis in #246
- Sprint19 plot variance (generalize to more geometries) by @amal-ghamdi in #214
- Sprint20 uq plot by @amal-ghamdi in #256
- Sprint20 update cont2d geom by @amal-ghamdi in #257
- Sprint20 enable nb in uq by @amal-ghamdi in #263
- fix output of model by @amal-ghamdi in #265
Full Changelog: v0.4.0...v0.5.0
v0.4.0
What's Changed
- Add Cauchy Distribution by @nabriis in #199
- Use scipy minimize solver for most problems solving for MAP by @nabriis in #201
- Rename Unadjusted Laplace Approximation sampler to UGLA by @nabriis in #217
- Add exact keyword to plot trace by @nabriis in #218
- fixed fail 2 issue number 181 by @Karl-JT in #220
- Update printing by @amal-ghamdi in #222
- Add links to plugins in CUQIpy documentation by @amal-ghamdi in #221
- Add Jacobian matrix option to Model by @nabriis in #216
- Rename diff distributions to MRF by @nabriis in #229
- Rename _1D naming scheme to just 1D by @nabriis in #240
- Rename MetropolisHasting to MH by @nabriis in #230
New Contributors
Full Changelog: v0.3.0...v0.4.0