Cleanup.

Document change linearindexes -> linearindices.
Clean up comments and unused code.

Author: traktofon

README.md

@@ -74,12 +74,12 @@ Indexing via symbols is used for this, specifically symbols `:L`,`:LP`,`:l`,`:lp
 are all equivalent. For example, `d[:L]` returns the localpart of `d`
 while `d[:L]=v` sets `v` as the localpart of `d`.
 
-* `localindexes(a::DArray)` gives a tuple of the index ranges owned by the
+* `localindices(a::DArray)` gives a tuple of the index ranges owned by the
 local process.
 
 * `convert(Array, a::DArray)` brings all the data to the local process.
 
-Indexing a `DArray` (square brackets) with ranges of indexes always
+Indexing a `DArray` (square brackets) with ranges of indices always
 creates a `SubArray`, not copying any data.
 
 
@@ -225,7 +225,7 @@ Garbage Collection and DArrays
 ------------------------------
 
 When a DArray is constructed (typically on the master process), the returned DArray objects stores information on how the
-array is distributed, which procesor holds which indexes and so on. When the DArray object
+array is distributed, which procesor holds which indices and so on. When the DArray object
 on the master process is garbage collected, all particpating workers are notified and
 localparts of the DArray freed on each worker.
 

src/darray.jl

@@ -198,7 +198,7 @@ function DArray(refs)
     nindices = Array{NTuple{length(dimdist),UnitRange{Int}}}(undef, dimdist...)
 
     for i in 1:length(nindices)
-        subidx = CartesianIndices(dimdist)[i] #ind2sub(dimdist, i)
+        subidx = CartesianIndices(dimdist)[i]
         nindices[i] = ntuple(length(subidx)) do x
             idx_in_dim = subidx[x]
             startidx = 1
@@ -525,7 +525,7 @@ Base.reshape(A::DArray{T,1,S}, d::Dims) where {T,S<:Array} = begin
         sztail = size(B)[2:end]
 
         for i=1:div(length(B),nr)
-            i2 = CartesianIndices(sztail)[i] #ind2sub(sztail, i)
+            i2 = CartesianIndices(sztail)[i]
             globalidx = [ I[j][i2[j-1]] for j=2:nd ]
 
             a = sub2ind(d, d1offs, globalidx...)

src/mapreduce.jl

@@ -84,17 +84,6 @@ Base.reducedim_initarray(A::DArray, region, v0, ::Type{R}) where {R} = begin
 end
 Base.reducedim_initarray(A::DArray, region, v0::T) where {T} = Base.reducedim_initarray(A, region, v0, T)
 
-#Base.reducedim_initarray0(A::DArray, region, v0, ::Type{R}) where {R} = begin
-#    # Store reduction on lowest pids
-#    pids = A.pids[ntuple(i -> i in region ? (1:1) : (:), ndims(A))...]
-#    chunks = similar(pids, Future)
-#    @sync for i in eachindex(pids)
-#        @async chunks[i...] = remotecall_wait(() -> Base.reducedim_initarray0(localpart(A), region, v0, R), pids[i...])
-#    end
-#    return DArray(chunks)
-#end
-#Base.reducedim_initarray0(A::DArray, region, v0::T) where {T} = Base.reducedim_initarray0(A, region, v0, T)
-
 # Compute mapreducedim of each localpart and store the result in a new DArray
 mapreducedim_within(f, op, A::DArray, region) = begin
     arraysize = [size(A)...]
@@ -144,7 +133,8 @@ end
 #end
 
 # TODO implement reduce, mapreduce with dims keyword
-# There are tests that call these on `DArray`s but they currently fall back to `AbstractArray` methods
+# There are tests that call these on `DArray`s but they currently run very slow
+# and I suspect that they fall back to `AbstractArray` methods
 
 
 function nnz(A::DArray)

test/runtests.jl

@@ -13,7 +13,6 @@ end
 @everywhere using DistributedArrays
 @everywhere using DistributedArrays.SPMD
 @everywhere using Random
-@everywhere using LinearAlgebra: LowerTriangular
 
 @everywhere srand(1234 + myid())