Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Remove a type restriction in Base.QuadGK.Segment (#19626) #19627

Merged
merged 4 commits into from
Dec 23, 2016
Merged
Show file tree
Hide file tree
Changes from 2 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion base/quadgk.jl
Original file line number Diff line number Diff line change
Expand Up @@ -40,7 +40,7 @@ immutable Segment
a::Number
b::Number
I
E::Real
E
end
isless(i::Segment, j::Segment) = isless(i.E, j.E)

Expand Down
96 changes: 96 additions & 0 deletions test/math.jl
Original file line number Diff line number Diff line change
Expand Up @@ -760,6 +760,102 @@ end
@test quadgk(cos, 0,0.7,1, norm=abs)[1] ≈ sin(1)
end

module Test19626
using Base.Test
# Begin tests for 19626: Unitful compatibility.
# Define a mock physical quantity type
immutable MockQuantity{T} <: Number
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It seems like it should work now even if it is not a subtype of Number?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If it is not a subtype of Number then I get a failure in vecnorm.

val::T
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is the <: Number necessary? I wouldn't bother parameterizing the type. Just make val::Float64.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

<: Number is necessary, otherwise I get an error from vecnorm.

end
Base.promote_rule{T,S}(::Type{MockQuantity{T}}, ::Type{MockQuantity{S}}) =
MockQuantity{promote_type(T,S)}
Base.convert{T}(::Type{MockQuantity{T}}, x::MockQuantity) = MockQuantity(T(x.val))

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You shouldn't need to define promotion rules and conversion for a minimal working example. Just define the minimum number of methods necessary for a test.

# Following definitions needed for quadgk to work with MockQuantity
import Base: +, -, *, abs, isnan, isinf, isless, <=
import Base.QuadGK: quadgk, do_quadgk
+(a::MockQuantity) = a
+(a::MockQuantity, b::MockQuantity) = MockQuantity(a.val+b.val)
-(a::MockQuantity) = MockQuantity(-a.val)
-(a::MockQuantity, b::MockQuantity) = MockQuantity(a.val-b.val)
*(a::MockQuantity, b::Number) = MockQuantity(a.val*b)
*(b::Number, a::MockQuantity) = MockQuantity(b*a.val)
abs(a::MockQuantity) = MockQuantity(abs(a.val))
isnan(a::MockQuantity) = isnan(a.val)
isinf(a::MockQuantity) = isinf(a.val)
isless(a::MockQuantity, b::MockQuantity) = isless(a.val, b.val)

# isless defn. necessary so that default abstol plays nicely with MockQuantity
isless(y::Number, x::MockQuantity) = y == 0 ? isless(MockQuantity(0), x) :
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I would just pass abstol = MockQuantity(0.0)

error("Dimensions issue")
isless(x::MockQuantity, y::Number) = y == 0 ? isless(x, MockQuantity(0)) :
error("Dimensions issue")

function quadgk{T}(f, a::MockQuantity{T}, b::MockQuantity{T},
c::MockQuantity{T}...; abstol=zero(T), reltol=sqrt(eps(T)),
maxevals=10^7, order=7, norm=vecnorm)

_do_quadgk(f, [a, b, c...], order, T, abstol, reltol, maxevals, norm)
end

# Necessary with infinite or semi-infinite intervals since !(MockQuantity <: Real)
# and do_quadgk tests if eltype(s) <: Real.
function _do_quadgk{Tw,T<:Real}(f, s::Array{MockQuantity{T},1}, n, ::Type{Tw},
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This seems overly complicated. For testing this PR, it should be sufficient to have an f(x) where the integrand f is a MockQuantity, but x is still Float64.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Agreed, I became overexuberant thinking about how I would do this in Unitful.

abstol, reltol, maxevals, nrm)

s_no_u = reinterpret(T, s)
s1 = s_no_u[1]; s2 = s_no_u[end]; inf1 = isinf(s1); inf2 = isinf(s2)
if inf1 || inf2
if inf1 && inf2 # x = t/(1-t^2) coordinate transformation
return do_quadgk(t -> begin t2 = t*t; den = 1 / (1 - t2);
# following line: f expects units; result should be multiplied by
# units of s (which we have stripped off)
MockQuantity(f(MockQuantity(t*den))) *
(1+t2)*den*den; end,
map(x -> isinf(x) ? copysign(one(x), x) :
2x / (1+hypot(1,2x)), s_no_u),
n, T, abstol, reltol, maxevals, nrm)
end
s0,si = inf1 ? (s2,s1) : (s1,s2)
if si < 0 # x = s0 - t/(1-t)
return do_quadgk(t -> begin den = 1 / (1 - t);
MockQuantity(f(MockQuantity(s0 - t*den))) *
den*den; end,
reverse!(map(x -> 1 / (1 + 1 / (s0 - x)), s_no_u)),
n, T, abstol, reltol, maxevals, nrm)
else # x = s0 + t/(1-t)
return do_quadgk(t -> begin den = 1 / (1 - t);
MockQuantity(f(MockQuantity(s0 + t*den))) *
den*den; end,
map(x -> 1 / (1 + 1 / (x - s0)), s_no_u),
n, T, abstol, reltol, maxevals, nrm)
end
end
do_quadgk(f, s, n, Tw, abstol, reltol, maxevals, nrm)
end

_do_quadgk{Tw}(f, s, n, ::Type{Tw}, abstol, reltol, maxevals, nrm) =
do_quadgk(f, s, n, Tw, abstol, reltol, maxevals, nrm)

# isapprox only needed for test purposes
Base.isapprox(a::MockQuantity, b::MockQuantity) = isapprox(a.val, b.val)

# Test physical quantity-valued functions
@test quadgk(x->MockQuantity(x), 0.0, 1.0)[1] ≈ MockQuantity(0.5)

# Test integration over an axis with units
@test quadgk(x->x.val, MockQuantity(0.0), MockQuantity(1.0))[1] ≈
MockQuantity(0.5)

# Test integration where the domain is infinite or semi-infinite
@test quadgk(x->exp(-x.val), MockQuantity(0.0), MockQuantity(Inf))[1] ≈
MockQuantity(1.0)
@test quadgk(x->exp(x.val), MockQuantity(-Inf), MockQuantity(0.0))[1] ≈
MockQuantity(1.0)
@test quadgk(x->exp(-abs(x.val)), MockQuantity(-Inf), MockQuantity(Inf))[1] ≈
MockQuantity(2.0)
end

@testset "subnormal flags" begin
# Ensure subnormal flags functions don't segfault
@test any(set_zero_subnormals(true) .== [false,true])
Expand Down