add dims keyword for padded_views and sym_paddedviews

[ashwanirathee]

@johnnychen94 Initial PR on addition of keyword dims in padded views and also in sym_padded_views later

closes #48

[johnnychen94]

Generally looks good, except that the current implementation is 2D specific; we actually want to support arbitrary dimensions.

I suggest that you write down some tests in test/runtest.jl and test it via

julia --project=. -e "using Pkg; Pkg.test()"

This helps you become confident in your implementation.

Also, when the functionality is done, don't forget to add documentation, and update README.md.

[johnnychen94]

To be clear, PaddedView handles arrays of arbitrary dimensions, for example, As can be a list of arrays with dimension N=4. I don't think the current code handles N>2 case.

[ashwanirathee]

@johnnychen94 Can the person provide the tuple in like in any random fashion..like can I provide it in following ways:

• (1,2,4,5,3) - I need sort and unique for this
• (1,2,3,4,5)- if it's continous and properly sorted,then it won't be a issue
• (1,1,1) - also this case
• (0,1,2,...) - also this case

[johnnychen94]

I'll follow the cat convention:

julia> cat([1,2], [3 4]; dims=(2,1))
3×3 Array{Int64,2}:
 1  0  0
 2  0  0
 0  3  4

[ashwanirathee]

Also we get dimension mismatch issue when arrays data is returned,which is mosaicviews data handling issue but that does arise for some reason I don't get yet...though we won't be talking about it here,right?

[johnnychen94]

The dimension mismatch issue you mentioned in #48 is not very clear to me, they should be unrelated. My best guess is that once we consolidate this feature in the PaddedViews side, the issue got fixed automatically.

FYI, there's a PR (JuliaArrays/MosaicViews.jl#31) related to the _padded_cat you're referring to, so no, I'm not going to check it until we get the current PaddedViews PR done.

[ashwanirathee]

Maybe the JuliaFormatter messed it all up a little,I formatted the whole code and now we have conflicts 😓 .I formatted it and it itself formatted too the code in vs code whenever save button is pushed.

[johnnychen94]

Yeah... could you revert the unrelated changes? It's a little bit mess to review the diffs now...

[ashwanirathee]

I think this does the job for the paddedviews, next is sym_ paddedviews

[johnnychen94]

Didn't check very carefully, but it looks good to me so far. Will recheck it when sym_paddedviews is supported.

[ashwanirathee]

without collect, it throws errors, Needs to check why

[johnnychen94]

without collect, it throws errors, Needs to check why

PaddedArray might not be 1-based Array, to properly compare two arrays, you need to make sure their axes are the same.

# from `@less ==(rand(4, 4), rand(4, 4))
function (==)(A::AbstractArray, B::AbstractArray)
    if axes(A) != axes(B)
        return false
    end
    anymissing = false
    for (a, b) in zip(A, B)
        eq = (a == b)
        if ismissing(eq)
            anymissing = true
        elseif !eq
            return false
        end
    end
    return anymissing ? missing : true
end

[johnnychen94]

LGTM, can you also update the docstrings and README examples? When that's done I think it's okay to merge this.