integrate changes from StatsModels 0.7 (upcoming release) (#664)

* statsmodels 0.7 upgrade * sort out ranef implicit intercept * compat requirement for GLM * fix sorting order for zerocorr * format * show methods for zerocorr * add in test from docs * Minor version bump, NEWS * bump julia version in docs ci --------- Co-authored-by: Phillip Alday <[email protected]>
JuliaStats · Apr 10, 2023 · c68bb4b · c68bb4b · palday · Apr 10, 2023
1 parent 658aea1
commit c68bb4b
Show file tree

Hide file tree

Showing 5 changed files with 38 additions and 10 deletions.
diff --git a/.github/workflows/documenter.yml b/.github/workflows/documenter.yml
@@ -24,7 +24,7 @@ jobs:
       - uses: actions/checkout@v2
       - uses: julia-actions/setup-julia@v1
         with:
-          version: 1.6
+          version: 1.8
       - uses: julia-actions/julia-buildpkg@latest
       - uses: julia-actions/julia-docdeploy@latest
         env:

diff --git a/NEWS.md b/NEWS.md
@@ -1,5 +1,6 @@
 MixedModels v4.9.0 Release Notes
 ==============================
+* Support `StatsModels` 0.7, drop support for `StatsModels` 0.6. [#664]
 * Revise code in benchmarks to work with recent Julia and PkgBenchmark.jl [#667]
 * Julia minimum compat version raised to 1.8 because of BSplineKit [#665]
 
@@ -396,5 +397,6 @@ Package dependencies
 [#653]: https://github.com/JuliaStats/MixedModels.jl/issues/653
 [#657]: https://github.com/JuliaStats/MixedModels.jl/issues/657
 [#663]: https://github.com/JuliaStats/MixedModels.jl/issues/663
+[#664]: https://github.com/JuliaStats/MixedModels.jl/issues/664
 [#665]: https://github.com/JuliaStats/MixedModels.jl/issues/665
 [#667]: https://github.com/JuliaStats/MixedModels.jl/issues/667
diff --git a/Project.toml b/Project.toml
@@ -1,7 +1,7 @@
 name = "MixedModels"
 uuid = "ff71e718-51f3-5ec2-a782-8ffcbfa3c316"
 author = ["Phillip Alday <[email protected]>", "Douglas Bates <[email protected]>", "Jose Bayoan Santiago Calderon <[email protected]>"]
-version = "4.8.2"
+version = "4.9.0"
 
 [deps]
 Arrow = "69666777-d1a9-59fb-9406-91d4454c9d45"
@@ -31,7 +31,7 @@ Tables = "bd369af6-aec1-5ad0-b16a-f7cc5008161c"
 Arrow = "1, 2"
 DataAPI = "1"
 Distributions = "0.21, 0.22, 0.23, 0.24, 0.25"
-GLM = "1.5.1"
+GLM = "1.8.2"
 JSON3 = "1"
 LazyArtifacts = "1"
 NLopt = "0.5, 0.6"
@@ -42,7 +42,7 @@ StaticArrays = "0.11, 0.12, 1"
 StatsAPI = "1.5"
 StatsBase = "0.31, 0.32, 0.33"
 StatsFuns = "0.8, 0.9, 1"
-StatsModels = "0.6.23"
+StatsModels = "0.7"
 StructTypes = "1"
 Tables = "1"
 julia = "1.8"

diff --git a/src/randomeffectsterm.jl b/src/randomeffectsterm.jl
@@ -37,7 +37,7 @@ function StatsModels.apply_schema(
     schema::MultiSchema{StatsModels.FullRank},
     Mod::Type{<:MixedModel},
 )
-    lhs, rhs = t.args_parsed
+    lhs, rhs = t.args
 
     isempty(intersect(StatsModels.termvars(lhs), StatsModels.termvars(rhs))) ||
         throw(ArgumentError("Same variable appears on both sides of |"))
@@ -53,11 +53,21 @@ check_re_group_type(term::GROUPING_TYPE) = true
 check_re_group_type(terms::Tuple{Vararg{<:GROUPING_TYPE}}) = true
 check_re_group_type(x) = false
 
+_unprotect(x) = x
+for op in StatsModels.SPECIALS
+    @eval _unprotect(t::FunctionTerm{typeof($op)}) = t.f(_unprotect.(t.args)...)
+end
+
 # make a potentially untyped RandomEffectsTerm concrete
 function StatsModels.apply_schema(
     t::RandomEffectsTerm, schema::MultiSchema{StatsModels.FullRank}, Mod::Type{<:MixedModel}
 )
-    lhs, rhs = t.lhs, t.rhs
+    # we need to do this here because the implicit intercept dance has to happen
+    # _before_ we apply_schema, which is where :+ et al. are normally
+    # unprotected.  I tried to finagle a way around this (using yet another
+    # schema wrapper type) but it ends up creating way too many potential/actual
+    # method ambiguities to be a good idea.
+    lhs, rhs = _unprotect(t.lhs), t.rhs
 
     # get a schema that's specific for the grouping (RHS), creating one if needed
     schema = get!(schema.subs, rhs, StatsModels.FullRank(schema.base.schema))
@@ -134,11 +144,11 @@ Base.:/(a::AbstractTerm, b::AbstractTerm) = a + a & b
 function StatsModels.apply_schema(
     t::FunctionTerm{typeof(/)}, sch::StatsModels.FullRank, Mod::Type{<:MixedModel}
 )
-    if length(t.args_parsed) ≠ 2
+    if length(t.args) ≠ 2
         throw(ArgumentError("malformed nesting term: $t (Exactly two arguments required"))
     end
 
-    first, second = apply_schema.(t.args_parsed, Ref(sch), Mod)
+    first, second = apply_schema.(t.args, Ref(sch), Mod)
 
     if !(typeof(first) <: CategoricalTerm)
         throw(
@@ -188,7 +198,7 @@ end
 function StatsModels.apply_schema(
     t::FunctionTerm{typeof(fulldummy)}, sch::StatsModels.FullRank, Mod::Type{<:MixedModel}
 )
-    return fulldummy(apply_schema.(t.args_parsed, Ref(sch), Mod)...)
+    return fulldummy(apply_schema.(t.args, Ref(sch), Mod)...)
 end
 
 # specify zero correlation
@@ -207,11 +217,21 @@ zerocorr(x) = ZeroCorr(x)
 # for schema extraction (from runtime-created zerocorr)
 StatsModels.terms(t::ZeroCorr) = StatsModels.terms(t.term)
 StatsModels.termvars(t::ZeroCorr) = StatsModels.termvars(t.term)
+StatsModels.degree(t::ZeroCorr) = StatsModels.degree(t.term)
+# dirty rotten no good ugly hack: make sure zerocorr ranef terms sort appropriately
+# cf https://github.com/JuliaStats/StatsModels.jl/blob/41b025409af03c0e019591ac6e817b22efbb4e17/src/terms.jl#L421-L422
+StatsModels.degree(t::FunctionTerm{typeof(zerocorr)}) = StatsModels.degree(only(t.args))
+
+Base.show(io::IO, t::ZeroCorr) = Base.show(io, MIME"text/plain"(), t)
+function Base.show(io::IO, ::MIME"text/plain", t::ZeroCorr)
+    # ranefterms already show with parens
+    return print(io, "zerocorr", t.term)
+end
 
 function StatsModels.apply_schema(
     t::FunctionTerm{typeof(zerocorr)}, sch::MultiSchema, Mod::Type{<:MixedModel}
 )
-    return ZeroCorr(apply_schema(t.args_parsed..., sch, Mod))
+    return ZeroCorr(apply_schema(only(t.args), sch, Mod))
 end
 
 function StatsModels.apply_schema(t::ZeroCorr, sch::MultiSchema, Mod::Type{<:MixedModel})

diff --git a/test/FactorReTerm.jl b/test/FactorReTerm.jl
@@ -157,6 +157,12 @@ end
         @test zc.rhs.sym == :subj
     end
 
+    @testset "Amalgamation of ZeroCorr with other terms" begin
+        f = @formula(reaction ~ 1 + days + (1|subj) + zerocorr(days|subj))
+        m = LMM(f, dataset(:sleepstudy), contrasts = Dict(:days => DummyCoding()))
+        re = only(m.reterms)
+        @test length(re.cnames) == length(unique(re.cnames)) == 10
+    end
 end
 
 @testset "Categorical Blocking Variable" begin