Merge pull request #8 from piever/coverage

Bring coverage to 100%

src/cox.jl

@@ -100,8 +100,12 @@ function StatsBase.coeftable(obj::CoxModel)
     ["Estimate", "Std.Error", "z value", "Pr(>|z|)"], ["x$i" for i in 1:length(β)], 4)
 end
 
-function Base.show(io::IO, obj::CoxModel)
-    println(io, "$(typeof(obj)):\n\nCoefficients:\n", coeftable(obj))
+function Base.show(io::IO, model::CoxModel)
+    ct = coeftable(model)
+    println(io, "$(typeof(model))")
+    println(io)
+    println(io,"Coefficients:")
+    show(io, ct)
 end
 
 StatsBase.coef(obj::CoxModel) = obj.β

test/runtests.jl

@@ -183,13 +183,32 @@ end
 
 
 @testset "Cox" begin
-    filepath = joinpath(@__DIR__, "rossi.csv")
+    filepath = joinpath(@__DIR__, "data", "rossi.csv")
     rossi = CSV.read(filepath)
     rossi[:event] = EventTime.(rossi[:week],rossi[:arrest] .== 1)
 
     outcome = coxph(@formula(event ~ fin+age+race+wexp+mar+paro+prio), rossi; tol = 1e-8)
     outcome_coefmat = coeftable(outcome)
 
+    regressor_matrix = Array(rossi[[:fin, :age, :race, :wexp, :mar, :paro, :prio]])
+    event_vector = rossi[:event]
+
+    outcome_without_formula = coxph(regressor_matrix, event_vector)
+
+    @test sprint(show, outcome_without_formula) == """
+        Survival.CoxModel{Float64}
+
+        Coefficients:
+               Estimate Std.Error   z value Pr(>|z|)
+        x1    -0.379416  0.191379  -1.98253   0.0474
+        x2   -0.0574299 0.0219988  -2.61059   0.0090
+        x3      0.31392  0.307995   1.01924   0.3081
+        x4     -0.14981  0.212226 -0.705898   0.4803
+        x5    -0.433724   0.38187  -1.13579   0.2560
+        x6   -0.0848615  0.195756 -0.433505   0.6646
+        x7     0.091521 0.0286469    3.1948   0.0014
+        """
+
     coef_matrix = ModelMatrix(ModelFrame(@formula(event ~ 0+fin+age+race+wexp+mar+paro+prio), rossi)).m
     outcome_from_matrix     = coxph(coef_matrix, rossi[:event]; tol = 1e-8, l2_cost = 0)
     outcome_from_matrix32   = coxph(Float32.(coef_matrix), rossi[:event]; tol = 1e-5)