Merge branch 'master' of https://github.com/FredericWantiez/AdvancedPS.jl into feature/traced_rng

Author: FredericWantiez

.JuliaFormatter.toml

@@ -0,0 +1 @@
+style="blue"

.github/workflows/Format.yml

@@ -0,0 +1,28 @@
+name: Format
+
+on:
+  push:
+    branches:
+      - master
+  pull_request:
+
+jobs:
+  format:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+      - uses: julia-actions/setup-julia@latest
+        with:
+          version: 1
+      - name: Format code
+        run: |
+          using Pkg
+          Pkg.add(; name="JuliaFormatter", uuid="98e50ef6-434e-11e9-1051-2b60c6c9e899")
+          using JuliaFormatter
+          format("."; verbose=true)
+        shell: julia --color=yes {0}
+      - uses: reviewdog/action-suggester@v1
+        if: github.event_name == 'pull_request'
+        with:
+          tool_name: JuliaFormatter
+          fail_on_error: true

Project.toml

@@ -1,7 +1,7 @@
 name = "AdvancedPS"
 uuid = "576499cb-2369-40b2-a588-c64705576edc"
 authors = ["TuringLang"]
-version = "0.2.0"
+version = "0.2.2"
 
 [deps]
 AbstractMCMC = "80f14c24-f653-4e6a-9b94-39d6b0f70001"
@@ -11,8 +11,8 @@ Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 StatsFuns = "4c63d2b9-4356-54db-8cca-17b64c39e42c"
 
 [compat]
-AbstractMCMC = "2"
-Distributions = "0.23, 0.24"
+AbstractMCMC = "2, 3"
+Distributions = "0.23, 0.24, 0.25"
 Libtask = "0.5"
 StatsFuns = "0.9"
 julia = "1.3"

README.md

@@ -3,6 +3,7 @@
 [![Build Status](https://github.com/TuringLang/AdvancedPS.jl/workflows/CI/badge.svg?branch=master)](https://github.com/TuringLang/AdvancedPS.jl/actions?query=workflow%3ACI%20branch%3Amaster)
 [![Coverage](https://codecov.io/gh/TuringLang/AdvancedPS.jl/branch/master/graph/badge.svg)](https://codecov.io/gh/TuringLang/AdvancedPS.jl)
 [![Coverage](https://coveralls.io/repos/github/TuringLang/AdvancedPS.jl/badge.svg?branch=master)](https://coveralls.io/github/TuringLang/AdvancedPS.jl?branch=master)
+[![Code Style: Blue](https://img.shields.io/badge/code%20style-blue-4495d1.svg)](https://github.com/invenia/BlueStyle)
 
 ### Reference
 

src/AdvancedPS.jl

@@ -1,10 +1,10 @@
 module AdvancedPS
 
-import AbstractMCMC
-import Distributions
-import Libtask
-import Random
-import StatsFuns
+using AbstractMCMC: AbstractMCMC
+using Distributions: Distributions
+using Libtask: Libtask
+using Random: Random
+using StatsFuns: StatsFuns
 
 include("resampling.jl")
 include("rng.jl")

src/container.jl

@@ -43,7 +43,7 @@ reset_model(f) = deepcopy(f)
 delete_retained!(f) = nothing
 
 # Task copying version of fork for Trace.
-function fork(trace::Trace, isref::Bool = false)
+function fork(trace::Trace, isref::Bool=false)
     newtrace = copy(trace)
     isref && delete_retained!(newtrace.f)
 
@@ -58,7 +58,7 @@ end
 function forkr(trace::Trace)
     newf = reset_model(trace.f)
 
-    ctask = let f=trace.ctask.task.code
+    ctask = let f = trace.ctask.task.code
         Libtask.CTask() do
             res = f()(trace.rng)
             Libtask.produce(nothing)
@@ -115,7 +115,7 @@ end
 function Base.push!(pc::ParticleContainer, p::Particle)
     push!(pc.vals, p)
     push!(pc.logWs, 0.0)
-    pc
+    return pc
 end
 
 # clones a theta-particle
@@ -126,7 +126,7 @@ function Base.copy(pc::ParticleContainer)
     # copy weights
     logWs = copy(pc.logWs)
 
-    ParticleContainer(vals, logWs)
+    return ParticleContainer(vals, logWs)
 end
 
 """
@@ -193,9 +193,9 @@ of the particle `weights`. For Particle Gibbs sampling, one can provide a refere
 function resample_propagate!(
     rng::Random.AbstractRNG,
     pc::ParticleContainer,
-    randcat = resample_systematic,
-    ref::Union{Particle, Nothing} = nothing;
-    weights = getweights(pc)
+    randcat=resample_systematic,
+    ref::Union{Particle,Nothing}=nothing;
+    weights=getweights(pc),
 )
     # check that weights are not NaN
     @assert !any(isnan, weights)
@@ -242,24 +242,24 @@ function resample_propagate!(
     pc.vals = children
     reset_logweights!(pc)
 
-    pc
+    return pc
 end
 
 function resample_propagate!(
     rng::Random.AbstractRNG,
     pc::ParticleContainer,
     resampler::ResampleWithESSThreshold,
-    ref::Union{Particle,Nothing} = nothing;
-    weights = getweights(pc)
+    ref::Union{Particle,Nothing}=nothing;
+    weights=getweights(pc),
 )
     # Compute the effective sample size ``1 / ∑ wᵢ²`` with normalized weights ``wᵢ``
     ess = inv(sum(abs2, weights))
 
     if ess ≤ resampler.threshold * length(pc)
-        resample_propagate!(rng, pc, resampler.resampler, ref; weights = weights)
+        resample_propagate!(rng, pc, resampler.resampler, ref; weights=weights)
     end
 
-    pc
+    return pc
 end
 
 """
@@ -300,9 +300,13 @@ function reweight!(pc::ParticleContainer)
 
     # The posterior for models with random number of observations is not well-defined.
     if numdone != 0
-        error("mis-aligned execution traces: # particles = ", n,
-              " # completed trajectories = ", numdone,
-              ". Please make sure the number of observations is NOT random.")
+        error(
+            "mis-aligned execution traces: # particles = ",
+            n,
+            " # completed trajectories = ",
+            numdone,
+            ". Please make sure the number of observations is NOT random.",
+        )
     end
 
     return false