Update svd.jl

[alanedelman]

No description provided.

[kshyatt]

Did you run these doctests locally? I'm pretty sure you wanted A == ... here...

[KristofferC]

I think this works fine? Note that it is ≈ (isapprox) and not =. However, there should be spaces between the inputs like

julia> A = rand(4,3);

julia> U,S,V = svd(A);

julia> A ≈ U*Diagonal(S)*V'
true

[kshyatt]

The dangers of phone reviewing...

…

On Dec 3, 2018, at 10:08 AM, Kristoffer Carlsson ***@***.***> wrote: I think this works fine? Note that it is ≈ (isapprox) and not =. However, there should be spaces between the inputs like julia> A = rand(4,3); julia> U,S,V = svd(A); julia> A ≈ U*Diagonal(S)*V' true — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

[simonbyrne]

Welcome @alanedelman, I see that this is your first pull request to Julia!

Would it be better to combine this example with the one below it?

[ViralBShah]

Tests failing with trailing whitespaces.

[andreasnoack]

I've fixed the doctests so they are now passing (the failure on Mac is unrelated). @alanedelman I've reorganized the paragraphs in the generalized svd documentation. The docstrings usually starts with a fairly compact description of inputs and outputs and then gives more details in the following. Please take a look.

I've also deleted the examples in the docstrings of the mutating versions and instead made references to the documentation of the non-mutating versions. This should make it easier to maintain the documentation.