Add 'eigmin'/'eigmax' methods for 'Eigen'

We add the methods 'eigmin' and 'eigmax' on the type 'F::Union{Eigen, GeneralizedEigen}' similarly to how 'eigvals' is defined on 'F'. Note that the corresponding unit tests in 'test/eigen.jl' that call 'eigvals' on such a struct are updated to include 'eigmin' and 'eigmax' calls as well.

(Resolves #1464 by @longemen3000.)

Author: Luis-Varona

src/eigen.jl

@@ -276,6 +276,8 @@ eigvecs(A::Union{Number, AbstractMatrix}; kws...) =
 eigvecs(F::Union{Eigen, GeneralizedEigen}) = F.vectors
 
 eigvals(F::Union{Eigen, GeneralizedEigen}) = F.values
+eigmin(F::Union{Eigen, GeneralizedEigen}) = minimum(F.values)
+eigmax(F::Union{Eigen, GeneralizedEigen}) = maximum(F.values)
 
 """
     eigvals!(A; permute::Bool=true, scale::Bool=true, sortby) -> values

test/eigen.jl

@@ -44,6 +44,8 @@ aimg  = randn(n,n)/2
             @test inv(a) ≈ inv(f)
             @test isposdef(a) == isposdef(f)
             @test eigvals(f) === f.values
+            @test eigmin(f) == minimum(f.values)
+            @test eigmax(f) == maximum(f.values)
             @test eigvecs(f) === f.vectors
             @test Array(f) ≈ a
 
@@ -81,6 +83,8 @@ aimg  = randn(n,n)/2
             @test prod(f.values) ≈ prod(eigvals(asym_sg/(ASG2))) atol=200ε
             @test eigvecs(asym_sg, ASG2) == f.vectors
             @test eigvals(f) === f.values
+            @test eigmin(f) == minimum(f.values)
+            @test eigmax(f) == maximum(f.values)
             @test eigvecs(f) === f.vectors
             @test_throws FieldError f.Z