Adjust definition of null model when there is no intercept (#14)

Technically this is breaking, but the current definition gives negative R² for models without the intercept.

Project.toml

@@ -1,7 +1,7 @@
 name = "StatsAPI"
 uuid = "82ae8749-77ed-4fe6-ae5f-f523153014b0"
 authors = ["Milan Bouchet-Valat <[email protected]"]
-version = "1.3.0"
+version = "1.4.0"
 
 [deps]
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"

src/statisticalmodel.jl

@@ -57,7 +57,11 @@ function islinear end
 """
     nulldeviance(model::StatisticalModel)
 
-Return the deviance of the null model, that is the one including only the intercept.
+Return the deviance of the null model, obtained by dropping all
+independent variables present in `model`.
+
+If `model` includes an intercept, the null model is the one with only the intercept;
+otherwise, it is the one without any predictor (not even the intercept).
 """
 function nulldeviance end
 
@@ -81,8 +85,11 @@ function loglikelihood end
 """
     nullloglikelihood(model::StatisticalModel)
 
-Return the log-likelihood of the null model corresponding to `model`.
-This is usually the model containing only the intercept.
+Return the log-likelihood of the null model, obtained by dropping all
+independent variables present in `model`.
+
+If `model` includes an intercept, the null model is the one with only the intercept;
+otherwise, it is the one without any predictor (not even the intercept).
 """
 function nullloglikelihood end