dataview to forward row and col views

[jishnub]

This adds methods to dataview to forward row and column views of BandedMatrices to the underlying data.

julia> B = BandedMatrix(0=>2:4, 1=>1:2)
3×3 BandedMatrix{Int64} with bandwidths (0, 1):
 2  1  ⋅
 ⋅  3  2
 ⋅  ⋅  4

julia> v = view(B, 1:2, 2)
2-element view(::BandedMatrix{Int64, Matrix{Int64}, Base.OneTo{Int64}}, 1:2, 2) with eltype Int64:
 1
 3

julia> BandedMatrices.dataview(v)
2-element view(::Vector{Int64}, 3:4) with eltype Int64:
 1
 3

Such methods would improve performance when indexing into the view, especially in loops.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 89.83%. Comparing base (c4e85f4) to head (fde3a83).
Report is 7 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #477      +/-   ##
==========================================
+ Coverage   89.71%   89.83%   +0.12%     
==========================================
  Files          25       27       +2     
  Lines        3685     3729      +44     
==========================================
+ Hits         3306     3350      +44     
  Misses        379      379              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[dlfivefifty]

Isn't this the same thing as bandeddata?

[jishnub]

Indeed, I had forgotten that the functionality already exists. This currently doesn't seem to work, though, but perhaps may be made to work:

julia> B = brand(4, 4, -1, 2)
4×4 BandedMatrix{Float64} with bandwidths (-1, 2):
  ⋅   0.326354  0.155544   ⋅ 
  ⋅    ⋅        0.856614  0.63237
  ⋅    ⋅         ⋅        0.483055
  ⋅    ⋅         ⋅         ⋅ 

julia> v = view(B, 1:2, 3)
2-element view(::BandedMatrix{Float64, Matrix{Float64}, Base.OneTo{Int64}}, 1:2, 3) with eltype Float64:
 0.1555437647782112
 0.8566140453180642

julia> BandedMatrices.bandeddata(v)
ERROR: MethodError: no method matching bandshift(::UnitRange{Int64}, ::Int64)
The function `bandshift` exists, but no method is defined for this combination of argument types.

Closest candidates are:
  bandshift(::Any)
   @ BandedMatrices ~/Dropbox/JuliaPackages/BandedMatrices.jl/src/banded/BandedMatrix.jl:880
  bandshift(::AbstractRange, ::Base.Slice{Base.OneTo{Int64}})
   @ BandedMatrices ~/Dropbox/JuliaPackages/BandedMatrices.jl/src/banded/BandedMatrix.jl:878
  bandshift(::AbstractRange, ::AbstractRange)
   @ BandedMatrices ~/Dropbox/JuliaPackages/BandedMatrices.jl/src/banded/BandedMatrix.jl:876
  ...

Stacktrace:
 [1] bandshift(S::SubArray{Float64, 1, BandedMatrix{Float64, Matrix{…}, Base.OneTo{…}}, Tuple{UnitRange{…}, Int64}, false})
   @ BandedMatrices ~/Dropbox/JuliaPackages/BandedMatrices.jl/src/banded/BandedMatrix.jl:880
 [2] bandeddata(V::SubArray{Float64, 1, BandedMatrix{Float64, Matrix{…}, Base.OneTo{…}}, Tuple{UnitRange{…}, Int64}, false})
   @ BandedMatrices ~/Dropbox/JuliaPackages/BandedMatrices.jl/src/banded/BandedMatrix.jl:917
 [3] top-level scope
   @ REPL[7]:1
Some type information was truncated. Use `show(err)` to see complete types.