Compat + Version bump (#53)

* version + compat bumps

* maxlog for some things

.github/workflows/ci.yml

@@ -18,7 +18,7 @@ jobs:
       fail-fast: false
       matrix:
         version:
-          - '1.4'
+          # - '1.6'
           - '1'
         R:
           - '4.0'

Project.toml

@@ -1,7 +1,7 @@
 name = "JellyMe4"
 uuid = "19ac8677-9a15-4623-9afd-84acc6165ce7"
 authors = ["Phillip Alday <[email protected]>"]
-version = "0.2.4"
+version = "0.2.5"
 
 [deps]
 CategoricalArrays = "324d7699-5711-5eae-9e2f-1d82baa6b597"
@@ -25,7 +25,7 @@ RCall = "0.13"
 StatsBase = "0.31, 0.32, 0.33"
 StatsModels = "0.6.8"
 Tables = "0.2, 1.0"
-julia = "1.4"
+julia = "1.6"
 
 [extras]
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"

src/glmerMod.jl

@@ -229,9 +229,9 @@ function sexp(::Type{RClass{:glmerMod}}, x::Tuple{GeneralizedLinearMixedModel{T}
     @debug r
     @warn """Some accuracy is lost in translation for GLMMs. This is fine with plotting, but proceed with caution for inferences.
              You can try letting the model "reconverge" in lme4 with
-                 update(model, control=glmerControl(optimizer="nloptwrap", calc.derivs=FALSE))."""
-    @info "lme4 handles deviance for GLMMs differently than MixedModels.jl"
-    @info "for the correct comparison, examine -2*logLik(RModel) and deviance(JuliaModel)"
+                 update(model, control=glmerControl(optimizer="nloptwrap", calc.derivs=FALSE)).""" maxlog=1
+    @info "lme4 handles deviance for GLMMs differently than MixedModels.jl" maxlog=1
+    @info "for the correct comparison, examine -2*logLik(RModel) and deviance(JuliaModel)" maxlog=1
     r = reval(r)
     r = protect(sexp(r))
     unprotect(1)

src/utilities.jl

@@ -53,8 +53,6 @@ this of course has a side effect: it changes the R dataframe
 function set_r_contrasts!(rdfname, formula)
     fixefform = first(formula.rhs)
 
-    warned_on_contrasts = false
-
     for tt in fixefform.terms
         if typeof(tt) <: CategoricalTerm
             R"""
@@ -73,11 +71,8 @@ function set_r_contrasts!(rdfname, formula)
             # then this should be handled by the coding of the first-order terms
             # if you are doing something that crazy, then you know enough linear algebra
             # to copy and interpret the relevant matrices directly
-            if !warned_on_contrasts
-                @info "contrasts on interaction terms are assumed to decompose into products of contrasts on the first-order terms"
-                @info "(if you don't know what that means, you're probably fine)"
-                warned_on_contrasts = true
-            end
+            @info "contrasts on interaction terms are assumed to decompose into products of contrasts on the first-order terms" maxlog=1
+            @info "(if you don't know what that means, you're probably fine)" maxlog=1
         end
     end
     rdfname