Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Space transformations refactor #449

Merged
merged 19 commits into from
Oct 19, 2024
Merged

Space transformations refactor #449

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
19 commits
Select commit Hold shift + click to select a range
783b7a4
Rename Iterator -> State, Tuner -> Tuning, restructure files
Micki-D Aug 30, 2024
e7ef101
Stash changes
Micki-D Sep 2, 2024
fb2c606
Introduce space transformations for MH MCMC samplers
Micki-D Sep 24, 2024
19e3c38
Clean up redundant code
Micki-D Sep 24, 2024
3eec0b9
Remove more redundant code, some renaming
Micki-D Sep 25, 2024
d711fd2
Reactivate HMC tests
Micki-D Sep 25, 2024
a084518
Rename mcmc_iterate! -> mcmc_iterate!!
Micki-D Sep 26, 2024
f2131ac
Introduce Super struct 'MCMCState' for MCMC sampling states, refactor…
Micki-D Sep 27, 2024
d87ecdf
Include updating of z-position for after tuning of mcmc samplers
Micki-D Sep 29, 2024
b271e28
Fix docs examples
Micki-D Sep 29, 2024
a66a552
Add docstrings for MCMC sampler types
Micki-D Sep 29, 2024
005956e
Add z-position update after post cycle tuning
Micki-D Sep 30, 2024
83f3a7e
Fix z-position update issue
Micki-D Oct 16, 2024
7833959
Set new types as experimental in docs
Micki-D Oct 16, 2024
4cf3048
Add MCMCChainState to experimental API docs
Micki-D Oct 16, 2024
95eb7a1
Add simple docstring for MCMCChainState
Micki-D Oct 17, 2024
ee4a89b
Remove obsolete code, rename sampling -> samplingalg in args/variables
Micki-D Oct 18, 2024
5e1ce8a
Use bat_default() in MCMCSampling
Micki-D Oct 19, 2024
eb950ea
Adjust new API in paper_example.jl
Micki-D Oct 19, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 10 additions & 0 deletions docs/src/experimental_api.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,4 +35,14 @@ ReactiveNestedSampling
SobolSampler
truncate_batmeasure
ValueAndThreshold

BAT.MCMCIterator
BAT.MCMCTunerState
BAT.TemperingState
BAT.MCMCProposalState
BAT.MCMCTempering
BAT.MCMCState
BAT.MCMCChainState
BAT.MCMCChainStateInfo
BAT.MCMCProposal
```
1 change: 0 additions & 1 deletion docs/src/internal_api.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -55,7 +55,6 @@ BAT.FullMeasureTransform
BAT.LFDensity
BAT.LFDensityWithGrad
BAT.LogDVal
BAT.MCMCIterator
BAT.MCMCSampleGenerator
BAT.MeasureLike
BAT.NoWhitening
Expand Down
2 changes: 1 addition & 1 deletion docs/src/stable_api.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -87,7 +87,7 @@ MCMCInitAlgorithm
MCMCMultiCycleBurnin
MCMCNoOpTuning
MCMCSampling
MCMCTuningAlgorithm
MCMCTuning
MetropolisHastings
MHProposalDistTuning
ModeAsDefined
Expand Down
7 changes: 3 additions & 4 deletions docs/src/tutorial_lit.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -230,7 +230,7 @@ posterior = PosteriorMeasure(likelihood, prior)
# Now we can generate a set of MCMC samples via [`bat_sample`](@ref). We'll
# use 4 MCMC chains with 10^5 MC steps in each chain (after tuning/burn-in):

samples = bat_sample(posterior, MCMCSampling(mcalg = MetropolisHastings(), nsteps = 10^5, nchains = 4)).result
samples = bat_sample(posterior, MCMCSampling(proposal = MetropolisHastings(), nsteps = 10^5, nchains = 4)).result
#md nothing # hide
#nb nothing # hide

Expand Down Expand Up @@ -377,8 +377,7 @@ plot!(-4:0.01:4, x -> fit_function(true_par_values, x), color=4, label = "Truth"
# We'll sample using the The Metropolis-Hastings MCMC algorithm:

mcmcalgo = MetropolisHastings(
weighting = RepetitionWeighting(),
tuning = AdaptiveMHTuning()
weighting = RepetitionWeighting()
)

# BAT requires a counter-based random number generator (RNG), since it
Expand Down Expand Up @@ -414,7 +413,7 @@ convergence = BrooksGelmanConvergence()
samples = bat_sample(
posterior,
MCMCSampling(
mcalg = mcmcalgo,
proposal = mcmcalgo,
nchains = 4,
nsteps = 10^5,
init = init,
Expand Down
2 changes: 1 addition & 1 deletion examples/paper-example/paper_example.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -143,7 +143,7 @@ posterior_bkg_signal = PosteriorMeasure(SignalBkgLikelihood(summary_dataset_tabl
nchains = 4
nsteps = 10^5

algorithm = MCMCSampling(mcalg = HamiltonianMC(), nchains = nchains, nsteps = nsteps)
algorithm = MCMCSampling(proposal = HamiltonianMC(), nchains = nchains, nsteps = nsteps)

samples_bkg = bat_sample(posterior_bkg, algorithm).result
eval_bkg = EvaluatedMeasure(posterior_bkg, samples = samples_bkg)
Expand Down
17 changes: 9 additions & 8 deletions ext/BATAdvancedHMCExt.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,20 +14,21 @@ using HeterogeneousComputing, AutoDiffOperators
using BAT: MeasureLike, BATMeasure

using BAT: get_context, get_adselector, _NoADSelected
using BAT: getalgorithm, mcmc_target
using BAT: MCMCIterator, MCMCIteratorInfo, MCMCChainPoolInit, MCMCMultiCycleBurnin, AbstractMCMCTunerInstance
using BAT: AbstractTransformTarget
using BAT: RNGPartition, set_rng!
using BAT: mcmc_step!, nsamples, nsteps, samples_available, eff_acceptance_ratio
using BAT: get_samples!, get_mcmc_tuning, reset_rng_counters!
using BAT: tuning_init!, tuning_postinit!, tuning_reinit!, tuning_update!, tuning_finalize!, tuning_callback
using BAT: getproposal, mcmc_target
using BAT: MCMCChainState, HMCState, HamiltonianMC, HMCProposalState, MCMCChainStateInfo, MCMCChainPoolInit, MCMCMultiCycleBurnin, MCMCTunerState, NoMCMCTempering
using BAT: _current_sample_idx, _proposed_sample_idx, _cleanup_samples
using BAT: AbstractTransformTarget, TriangularAffineTransform
using BAT: RNGPartition, get_rng, set_rng!
using BAT: mcmc_step!!, nsamples, nsteps, samples_available, eff_acceptance_ratio
using BAT: get_samples!, reset_rng_counters!
using BAT: create_trafo_tuner_state, create_proposal_tuner_state, mcmc_tuning_init!!, mcmc_tuning_postinit!!, mcmc_tuning_reinit!!, mcmc_tune_transform_post_cycle!!, transform_mcmc_tuning_finalize!!, tuning_callback
using BAT: totalndof, measure_support, checked_logdensityof
using BAT: CURRENT_SAMPLE, PROPOSED_SAMPLE, INVALID_SAMPLE, ACCEPTED_SAMPLE, REJECTED_SAMPLE

using BAT: HamiltonianMC
using BAT: AHMCSampleID, AHMCSampleIDVector
using BAT: HMCMetric, DiagEuclideanMetric, UnitEuclideanMetric, DenseEuclideanMetric
using BAT: HMCTuningAlgorithm, MassMatrixAdaptor, StepSizeAdaptor, NaiveHMCTuning, StanHMCTuning
using BAT: HMCTuning, MassMatrixAdaptor, StepSizeAdaptor, NaiveHMCTuning, StanHMCTuning

using ValueShapes: varshape

Expand Down
Loading
Loading