added has_intercept (#47)

src/statsmodel.jl

@@ -45,12 +45,28 @@ struct DataFrameRegressionModel{M,T} <: RegressionModel
     mm::ModelMatrix{T}
 end
 
+"""
+    drop_intercept(::Type)
+
+Define whether a given model automatically drops the intercept. Return `false` by default. 
+To specify that a model type `T` drops the intercept, overload this function for the 
+corresponding type: `drop_intercept(::Type{T}) = true`
+
+Models that drop the intercept will be fitted without one: the intercept term will be 
+removed even if explicitly provided by the user. Categorical variables will be expanded 
+in the rank-reduced form (contrasts for `n` levels will only produce `n-1` columns).
+"""
+drop_intercept(::Type) = false
+
 for (modeltype, dfmodeltype) in ((:StatisticalModel, DataFrameStatisticalModel),
                                  (:RegressionModel, DataFrameRegressionModel))
     @eval begin
         function StatsBase.fit(::Type{T}, f::Formula, df::AbstractDataFrame,
                                args...; contrasts::Dict = Dict(), kwargs...) where T<:$modeltype
-            mf = ModelFrame(f, df, contrasts=contrasts)
+            trms = Terms(f)
+            drop_intercept(T) && (trms.intercept = true)
+            mf = ModelFrame(trms, df, contrasts=contrasts)
+            drop_intercept(T) && (mf.terms.intercept = false)
             mm = ModelMatrix(mf)
             y = model_response(mf)
             $dfmodeltype(fit(T, mm.m, y, args...; kwargs...), mf, mm)

test/statsmodel.jl

@@ -89,6 +89,41 @@ fit(DummyMod, f3, d, contrasts = Dict(:x1p => EffectsCoding(),
 @test_throws Exception fit(DummyMod, f3, d, contrasts = Dict(:x1p => EffectsCoding(),
                                                              :x2p => 1))
 
+# A dummy RegressionModel type that does not support intercept
+struct DummyModNoIntercept <: RegressionModel
+    beta::Vector{Float64}
+    x::Matrix
+    y::Vector
+end
+
+StatsModels.drop_intercept(::Type{DummyModNoIntercept}) = true
+
+## dumb fit method: just copy the x and y input over
+StatsBase.fit(::Type{DummyModNoIntercept}, x::Matrix, y::Vector) =
+    DummyModNoIntercept(collect(1:size(x, 2)), x, y)
+StatsBase.model_response(mod::DummyModNoIntercept) = mod.y
+## dumb coeftable: just prints the "beta" values
+StatsBase.coeftable(mod::DummyModNoIntercept) =
+    CoefTable(reshape(mod.beta, (size(mod.beta,1), 1)),
+             ["'beta' value"],
+             ["" for n in 1:size(mod.x,2)],
+             0)
+
+f1 = @formula(y ~ 1 + x1 * x2)
+f2 = @formula(y ~ 0 + x1 * x2)
+m1 = fit(DummyModNoIntercept, f1, d)
+m2 = fit(DummyModNoIntercept, f2, d)
+ct1 = coeftable(m1)
+ct2 = coeftable(m2)
+@test ct1.rownms == ct2.rownms == ["x1", "x2", "x1 & x2"]
+
+f1 = @formula(y ~ 1 + x1p)
+f2 = @formula(y ~ 0 + x1p)
+m1 = fit(DummyModNoIntercept, f1, d)
+m2 = fit(DummyModNoIntercept, f2, d)
+ct1 = coeftable(m1)
+ct2 = coeftable(m2)
+@test ct1.rownms == ct2.rownms == ["x1p: 6", "x1p: 7", "x1p: 8"]
 
 ## Another dummy model type to test fall-through show method
 struct DummyModTwo <: RegressionModel