dep from StatsFuns instead of Distributions (#195)

* dep from StatsFuns instead of Distributions

StatsModels should remain as ligthweight as possible.

`Distribution` depends on `StatsFuns`, but also on packages such as`StaticArrays` that create a lot of invalidation.
On my computer, this decreases `using StatsModels` by a third.

* Update Project.toml

Author: matthieugomez

Project.toml

@@ -1,26 +1,26 @@
 name = "StatsModels"
 uuid = "3eaba693-59b7-5ba5-a881-562e759f1c8d"
-version = "0.6.14"
+version = "0.6.15"
 
 [deps]
 DataAPI = "9a962f9c-6df0-11e9-0e5d-c546b8b5ee8a"
 DataStructures = "864edb3b-99cc-5e75-8d2d-829cb0a9cfe8"
-Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
 Printf = "de0858da-6303-5e67-8744-51eddeeeb8d7"
 ShiftedArrays = "1277b4bf-5013-50f5-be3d-901d8477a67a"
 SparseArrays = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"
 StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
+StatsFuns = "4c63d2b9-4356-54db-8cca-17b64c39e42c"
 Tables = "bd369af6-aec1-5ad0-b16a-f7cc5008161c"
 
 [compat]
 CategoricalArrays = "0.8"
 DataAPI = "1.1"
 DataFrames = "0.21"
 DataStructures = "0.17.0, 0.18"
-Distributions = "0.16, 0.17, 0.18, 0.19, 0.20, 0.21, 0.22, 0.23"
 ShiftedArrays = "1.0.0"
 StatsBase = "0.22, 0.23, 0.24, 0.25, 0.26, 0.27, 0.28, 0.29, 0.30, 0.31, 0.32, 0.33"
+StatsFuns = "0.9"
 Tables = "0.2, 1"
 julia = "1"
 

src/StatsModels.jl

@@ -8,7 +8,7 @@ using DataStructures
 using DataAPI
 using DataAPI: levels
 using Printf: @sprintf
-using Distributions: Chisq, ccdf
+using StatsFuns: chisqccdf
 
 using SparseArrays
 using LinearAlgebra

src/lrtest.jl

@@ -116,7 +116,7 @@ function lrtest(mods::StatisticalModel...; atol::Real=0.0)
                             "in models with more degrees of freedom"))
     end
 
-    pval = (NaN, ccdf.(Chisq.(abs.(Δdf)), abs.(Δdev))...)
+    pval = (NaN, chisqccdf.(abs.(Δdf), abs.(Δdev))...)
     return LRTestResult(Int(nobs(mods[1])), dev, df, pval)
 end