Make tests Julia 1.0 compatible

src/Romberg.jl

@@ -78,7 +78,7 @@ is a prime number, it is nearly equivalent to the trapezoidal rule without extra
 # Examples
 
 ```jldoctest
-julia> x = range(0, π, length=2^8+1);
+julia> x = range(0, stop=π, length=2^8+1);
 
 julia> romberg(x, sin.(x))
 (2.0000000000000018, 1.9984014443252818e-15)

test/runtests.jl

@@ -2,17 +2,17 @@ using Romberg, Test
 
 @testset "Romberg.jl" begin
     # Test interfaces
-    x = range(0, π, length=2^8+1)
+    x = range(0, stop=π, length=2^8+1)
     y = sin.(x)
 
     @test romberg(x, y) == romberg(x, y, maxeval=9)
 
     # Test ispow2(length(x) - 1) == false
-    x = range(0, π, length=2^8)
+    x = range(0, stop=π, length=2^8)
     y = sin.(x)
     @test romberg(x, y)[1] ≈ 2   rtol=1e-12
 
-    x = range(0, π, length=2^8-1)
+    x = range(0, stop=π, length=2^8-1)
     y = sin.(x)
     @test romberg(x, y)[1] ≈ 2   rtol=1e-9
 
@@ -22,31 +22,31 @@ using Romberg, Test
     @test romberg(x, y)[1] == 0
 
     # Test length(x) == 2
-    x = range(0, 1, length=2)
+    x = range(0, stop=1, length=2)
     y = x.^2
     @test romberg(x,y)[1] == 0.5   # integration is inaccurate
 
     # Test length(x) == 3
-    x = range(0, 1, length=3)
+    x = range(0, stop=1, length=3)
     y = x.^2
     @test romberg(x, y)[1] ≈ 1/3   rtol=3e-16  # should be exact up to roundoff error
 
     # Integrate different functions
-    x = range(0, π, length=2^8+1)
+    x = range(0, stop=π, length=2^8+1)
     y = sin.(x)
     @test romberg(x, y)[1] ≈ 2    rtol=1e-15
 
-    x = range(0, 1, length=2^8+1)
+    x = range(0, stop=1, length=2^8+1)
     y = x.^3
     @test romberg(x, y)[1] ≈ 1/4   rtol=3e-16  # should be exact up to roundoff error
 
-    x = range(0, π/2, length=2^5+1)
+    x = range(0, stop=π/2, length=2^5+1)
     y = sin.(x).^2
     @test romberg(x, y)[1] ≈ π/4    rtol=1e-15
 
     m = 3
     n = 4
-    x = range(0, π, length=2^8+1)
+    x = range(0, stop=π, length=2^8+1)
     y = sin.(m*x).*cos.(n*x)
     v = romberg(x, y)
     @test v[1] ≈ 2*m/(m^2 - n^2)  rtol=1e-13
@@ -57,18 +57,18 @@ using Romberg, Test
 
     # this one requires lots of samples...
     a = 15
-    x = range(0, a, length=2^16+1)
+    x = range(0, stop=a, length=2^16+1)
     y = sqrt.(a^2 .- x.^2)
     @test romberg(x, y)[1] ≈ π*a^2/4   rtol=1e-8
     # we can do much better if we put in the correct convergence
     # rate for this singular integrand, but this requires some
     # understanding of the theory that most people won't have…
-    x = range(0, a, length=2^7+1)
+    x = range(0, stop=a, length=2^7+1)
     y = sqrt.(a^2 .- x.^2)
     @test romberg(x, y, power=0.5)[1] ≈ π*a^2/4   rtol=1e-8
 
     # tricky to integrate
-    x = range(1e-15, 1, length=2^16+1)
+    x = range(1e-15, stop=1, length=2^16+1)
     y = log.(x)./(1 .+ x)
     @test romberg(x, y)[1] ≈ -π^2/12  atol=1e-4