Merge e332060 into 600b82c

Project.toml

@@ -16,8 +16,18 @@ ProgressLogging = "33c8b6b6-d38a-422a-b730-caa89a2f386c"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 Setfield = "efcf1570-3423-57d1-acb7-fd33fddbac46"
 
+[weakdeps]
+AdvancedHMC = "0bf59076-c3b1-5ca4-86bd-e02cd72cde3d"
+AdvancedMH = "5b7e9947-ddc0-4b3f-9b55-0d8042f74170"
+
+[extensions]
+MCMCTemperingAdvancedHMCExt = ["AdvancedHMC"]
+MCMCTemperingAdvancedMHExt = ["AdvancedMH"]
+
 [compat]
 AbstractMCMC = "5"
+AdvancedHMC = "0.6"
+AdvancedMH = "0.8"
 ConcreteStructs = "0.2"
 Distributions = "0.24, 0.25"
 DocStringExtensions = "0.8, 0.9"
@@ -27,3 +37,7 @@ MCMCChains = "5, 6"
 ProgressLogging = "0.1"
 Setfield = "0.7, 0.8, 1"
 julia = "1"
+
+[extras]
+AdvancedHMC = "0bf59076-c3b1-5ca4-86bd-e02cd72cde3d"
+AdvancedMH = "5b7e9947-ddc0-4b3f-9b55-0d8042f74170"

ext/MCMCTemperingAdvancedHMCExt.jl

@@ -0,0 +1,19 @@
+module MCMCTemperingAdvancedHMCExt
+
+using MCMCTempering: MCMCTempering, Setfield
+using AdvancedHMC: AdvancedHMC
+
+MCMCTempering.getparams_and_logprob(t::AdvancedHMC.Transition) = t.z.θ, t.z.ℓπ.value
+MCMCTempering.getparams_and_logprob(state::AdvancedHMC.HMCState) = MCMCTempering.getparams_and_logprob(state.transition)
+
+# TODO: Implement `state_from` instead, to avoid re-computation of gradients if possible.
+function MCMCTempering.setparams_and_logprob!!(model, state::AdvancedHMC.HMCState, params, lp)
+ # NOTE: Need to recompute the gradient because it might be used in the next integration step.
+ hamiltonian = AdvancedHMC.Hamiltonian(state.metric, model)
+ return Setfield.@set state.transition.z = AdvancedHMC.phasepoint(
+ hamiltonian, params, state.transition.z.r;
+ ℓκ=state.transition.z.ℓκ
+ )
+end
+
+end

test/compat.jl → ext/MCMCTemperingAdvancedMHExt.jl

@@ -1,4 +1,8 @@
-# AdvancedMH.jl
+module MCMCTemperingAdvancedMHExt
+
+using MCMCTempering: MCMCTempering, Setfield
+using AdvancedMH: AdvancedMH
+
 MCMCTempering.getparams_and_logprob(transition::AdvancedMH.Transition) = transition.params, transition.lp
 function MCMCTempering.setparams_and_logprob!!(transition::AdvancedMH.Transition, params, lp)
  Setfield.@set! transition.params = params
@@ -17,16 +21,4 @@ function MCMCTempering.setparams_and_logprob!!(model, transition::AdvancedMH.Gra
  )
 end
 
-# AdvancedHMC.jl
-MCMCTempering.getparams_and_logprob(t::AdvancedHMC.Transition) = t.z.θ, t.z.ℓπ.value
-MCMCTempering.getparams_and_logprob(state::AdvancedHMC.HMCState) = MCMCTempering.getparams_and_logprob(state.transition)
-
-# TODO: Implement `state_from` instead, to avoid re-computation of gradients if possible.
-function MCMCTempering.setparams_and_logprob!!(model, state::AdvancedHMC.HMCState, params, lp)
- # NOTE: Need to recompute the gradient because it might be used in the next integration step.
- hamiltonian = AdvancedHMC.Hamiltonian(state.metric, model)
- return Setfield.@set state.transition.z = AdvancedHMC.phasepoint(
- hamiltonian, params, state.transition.z.r;
- ℓκ=state.transition.z.ℓκ
- )
 end

test/setup.jl

@@ -19,4 +19,3 @@ using Turing: Turing, DynamicPPL
 
 include("utils.jl")
 include("test_utils.jl")
-include("compat.jl")