fix compat, bump version

.github/workflows/runtests.yml

@@ -15,15 +15,15 @@ jobs:
       matrix:
         version:
           - '1.6'
-          - '1.8'
+          - '1.9'
         os:
           - ubuntu-latest
           - macOS-latest
         arch:
           - x64
     steps:
-      - uses: actions/checkout@v3
-      - uses: julia-actions/setup-julia@v1.8
+      - uses: actions/checkout@v4
+      - uses: julia-actions/setup-julia@v1.9
         with:
           version: ${{ matrix.version }}
           arch: ${{ matrix.arch }}

Project.toml

@@ -1,7 +1,7 @@
 name = "EqualitySampler"
 uuid = "78feedfb-ee96-4c5d-801a-e07e5d165ffd"
-authors = ["Don van den Bergh <donvdbergh@hotmail.com> and Fabian Dablander"]
-version = "0.1.0"
+authors = ["Don van den Bergh <d.vandenbergh@uva.nl> and Fabian Dablander"]
+version = "0.1.1"
 
 [deps]
 AbstractMCMC = "80f14c24-f653-4e6a-9b94-39d6b0f70001"
@@ -32,26 +32,26 @@ Suppressor = "fd094767-a336-5f1f-9728-57cf17d0bbfb"
 Turing = "fce5fe82-541a-59a6-adf8-730c64b5f9a0"
 
 [compat]
-AbstractMCMC = "4"
-AdvancedHMC = "0.3, 0.4"
-Bijectors = "0.10"
+AbstractMCMC = "4, 5"
+AdvancedHMC = "0.3, 0.4, 0.5"
+Bijectors = "0.10, 0.11, 0.12, 0.13"
 Combinatorics = "1"
 DataFrames = "1.3"
 Distributions = "0.25"
 DistributionsAD = "0.6"
 DocStringExtensions = "0.8, 0.9"
-DynamicPPL = "0.19, 0.20, 0.21"
-FillArrays = "0.13"
-GLM = "1"
-KernelDensity = "0.6"
+DynamicPPL = "0.19, 0.20, 0.21, 0.22, 0.23, 0.24"
+FillArrays = "1"
+GLM = "1.8"
+KernelDensity = "0.6, 0.7"
 LogExpFunctions = "0.3"
-MCMCChains = "5"
+MCMCChains = "5, 6"
 OrderedCollections = "1.4"
 PDMats = "0.11"
 QuadGK = "2"
 SpecialFunctions = "2"
-StatsBase = "0.33"
-StatsModels = "0.6"
+StatsBase = "0.33, 0.34"
+StatsModels = "0.6, 0.7"
 Suppressor = "0.2"
-Turing = "0.21, 0.22, 0.23, 0.24"
+Turing = "0.21, 0.22, 0.23, 0.24, 0.25, 0.26, 0.27, 0.28, 0.29"
 julia = "1.6"

src/EqualitySampler.jl

@@ -20,41 +20,6 @@ using DocStringExtensions
 
 export
 
-	#=
-
-		TODO:
-
-			Work out package idea for efficient log likelihood evaluation through sufficient statistics
-
-				- work out abstract types, e.g., AbstractSuffStatDistribution
-				- Distribution over tuple?
-				- Example MvNormal
-				struct MvNormalSuffStat{D}
-					dist::D
-					params::typeof(params(D))
-					suffstats
-				end
-				# does this work with D as parametric type or does it need to be a field?
-				- rand fallback calls compute_suffstats(D, rand(D, n))
-
-
-	=#
-
-	#=
-		TODO:
-
-		- [ ] write and export one function that computes the likelihood using Distributions.suffstats(MvNormal, x) and Distributions.suffstats(Normal, x)
-			perhaps skip/ remove logpdf_mv_normal_chol_suffstat and friends (constructing the full matrix is not super expensive)
-			also use PDMats.invquad and friends
-
-		- [ ] consider renaming "model" to "partition" everywhere. For example,
-			AbstractPartitionDistribution	-> AbstractPartitionDistribution
-			reduce_model 				-> normalize_partition
-
-
-
-	=#
-
 	get_normal_dense_chol_suff_stats,
 	logpdf_mv_normal_chol_suffstat,
 	loglikelihood_suffstats,

test/test-anova_test.jl

@@ -1,3 +1,4 @@
+using EqualitySampler, Test
 import Random, Distributions, MCMCChains, AbstractMCMC, Statistics, Turing
 import EqualitySampler.Simulations