Activity for Generalized Approximate Message Passing
-
Generalized Approximate Message Passing released /gampmatlab20200328.zip
-
Philip Schniter committed [r633]
Added a new version of VAMP, VampSlmEst3.m, which has improved damping and can be run in ADMM-PR mode for initial iterations, as well as VampOptEst3.m
-
Fixed bug in EMBGAMP.m: needed to report history
-
updated binary
-
added an option to FistaOpt that lets fistaEst run without acceleration (i.e., ISTA)
-
Allow option "xvar0" in ampEst.m
-
Added support for complex-valued denoisers in FxnhandleEstimIn.m
-
main/GMEstimIn.m: automatically check that mixture weights sum to one
-
Added a prVAMP demo that estimates noise variance (from Chris Metzler).
-
Fixed a problem with xvar in main/TuneVarScaleEstimIn.m and added comparison to monte-carlo xvar estimate in test/testTuneVarScaleEstimIn.m
-
Added main/TuneVarScaleEstimIn.m, an EstimIn wrapper that estimates both an unknown scaling and variance on rhat.
-
Fixed problem in main/DisScaEstim.m where complex-valued inputs led to complex-valued variances
-
Changed syntax of warning in VampGlmEst*.m due to changes in Matlab 2018a
-
Changed syntax of warnings in VampSlmEst2.m due to
-
Added fxnErr3 to VampGlmEst & VampGlmEst2, which is an error-reporting function that has a very generic dependence on the various VampGlm states (as compared to fxnErr1 and fxnErr2, which both depend only on x1 and z1).
-
Minor changes to ncCAwgnEstimOut.m and VampGlmEst2.m for GPU compatibility
-
-
PBiGAMPProblem: Added a real constructor.
-
Updated demo_prVAMP.m to be compatibile with new VampGlmEst2.m
-
Added warm-starting capability to VampGlmEst2.m, requiring changes in VampGlmOpt.m.
-
Fixed Matlab compatibility bug in demo_prVAMP.m
-
disambig1Dfft.m: now works on multiple columns. demo_prVAMP.m: new damping scheme via VampGlmEst2, and various other small improvements.
-
VampGlmEst.m: general cleanup. VampGlmEst2.m: now allows damping of up to 12 different variables. VampGlmOpt.m: includes dampConfig variable. demoVampGlm.m: now includes phase-retrieval and various other tweaks.
-
Added the option to use custom error-reporting functions in gampEst.m (and GampOpt.m). Also, updated demo_prVAMP.m so that it works with sparse signals.
-
Added missing size() and/or numColumn() methods
-
Cleanup of GLM-VAMP Demo and State Evolution
-
Added "opt.altUpdate" option that alternates between update of x and z in VampGlmEst2.m
-
Added VampGlmEst2.m, which has an improved method of damping
-
Renaming of var as var1 in TuneVarEstimIn.m for interpretability.
-
Added TuneVarEstimIn.m for "adaptive VAMP/GAMP"
-
Avoided use of "size(A,dim)" fxn in VampSlmEst.m, VampSlmEst2.m, VampGlmEst.m
-
demo_prVAMP.m: oops again! fixed.
-
demo_prVAMP.m: oops! small bug
-
demo_prVAMP.m: a demo for phase-retrieval VAMP
-
VampGlmOpt.m: Reduced default value of gamMax f...
-
VampSlmEst2.m: added dampGam, which is a separa...
-
Added QuantizeEstimOut.m and CQuantizeEstimOout.m
-
Increased default variance of initialization in...
-
Revised group-sparse code for HyGAMP paper, mak...
-
GampGlmEst.m: added debug mode
-
-
Added "help" comments to VAMP/VampGlmEst.m.
-
VampSlmOpt.m: added "NoisePrecMax" option
-
demoVampGlm.m & demoVampSlm.m: switched to Fxnh...
-
VampGlmEst.m: initialization check for MMV case
-
ampSE.m: Added the ability to initialize mse.
-
VampSlmEst.m: Added |dr|/|r| reporting in verbo...
-
Flipped the unwanted executable bit on a bunch ...
-
main/ampEst.m: fixed scaling bug in wvar case o...
-
-
Added VAMP routines.
-
ampEst.m, AmpOpt.m: added rvarMethod option
-
ampEst.m & AmpOpt.m: implement standard undampe...
-
Added EstimInAvg.m, EstimOutAvg.m, and EstimOut...
-
stateEvo/EstimInAvg.m: generalization useful wh...
-
Added stateEvo/EstimInAvg.m to help with GEC st...
-
FxnhandleEstimIn.m: Improved divergence estimat...
-
FxnhandleEstimIn.m: use mean-squared rhat to se...
-
FxnhandleEstimIn.m: robust calculation of diver...
-
FxnhandleEstimIn.m: added optional input of dir...
-
Added some more example files to /examples/PBiGAMP
-
ProbitEstimOut.m: Added ML and EM autoTuning me...
-
SparseScaEstim.m: robustified xvar computation
-
ProbitEstimOut.m: added Size() method
-
(C)AwgnEstimOut.m: Fixed a small bug in EM-styl...
-
(C)AwgnEstimOut.m: removed warnings when settin...
-
SparseScaEstim.m: Changed superclass from handl...
-
SoftThreshDMMEstimIn.m: generalized to work in ...
-
AwgnEstimIn.m and CAwgnEstimIn.m: Fixed small b...
-
SoftThreshDMMEstimIn.m and SoftThreshEstimIn.m:...
-
Added a file to examples/ProbitSE that experime...
-
FxnhandleLinTrans.m: clarified scaling on Fro2 ...
-
Added a few lines to EMGM example/demo files to...
-
PBiGAMP.m: fixed a small bug where an error is ...
-
-
Jeremy Vila committed [r562]
Added EMGMAMPMOS functionality to work with upd...
-
(C)AwgnEstimOut.m: Added third output to .estim...
-
PBiGAMP: Added affine example
-
PBiGAMP(Opt).m: Added option parameter "errTun...
-
(C)AwgnEstimOut.m: Added two new tuning method...
-
PBiGAMPsimple.m: A wrapper around PBiGAMP that...
-
PBiGAMP.m / PBiGAMPOpt.m: added an option to br...
-
PBiGAMP.m / EMPBigAMP.m / PBiGAMPOpt.m: Moved ...
-
PBiGAMP:trial_affine.m and script_affine.m: Ext...
-
PBiGAMP/path_setup.m: Hopefully this file now w...
-
PBiGAMP.m: 1) Fixed bug where distCountB/C and ...
-
EMGMAMP: moved new stuff to directory "EMGMAMPn...
-
Refinements to EMGMAMP, EMBGAMP, EMOpt, check_o...
-
(C)AwgnEstimOut: Changed default value of noise...
-
EMGMAMP: now uses (C)AwgnEstimOut with internal...
-
(C)AwgnEstimOut: Now includes internal noise-va...
-
PBiGAMP: Minor bug fixes and a new option that ...
-
Added a new feature to FxnhandleLinTrans: user ...
-
Added several files for multiclass classification
-
Added a new output channel for multinomial logi...
-
Moved various classification-specific output ch...
-
-
Added missing file to PBiGAMP folder (sourced f...
-
Removed some unnecessary files.
-
Jason Parker committed [r539]
1 >
