deprecate diagm(A::SparseMatrixCSC) in favor of diagm(sparsevec(A))

Author: fredrikekre

base/deprecated.jl

@@ -1687,6 +1687,9 @@ export hex2num
 # issue #5148, PR #23259
 # warning for `const` on locals should be changed to an error in julia-syntax.scm
 
+# PR 23341
+@deprecate diagm(A::SparseMatrixCSC) diagm(sparsevec(A))
+
 # END 0.7 deprecations
 
 # BEGIN 1.0 deprecations

base/sparse/sparsematrix.jl

@@ -3413,49 +3413,6 @@ function trace(A::SparseMatrixCSC{Tv}) where Tv
     return s
 end
 
-function diagm(v::SparseMatrixCSC{Tv,Ti}) where {Tv,Ti}
-    if size(v,1) != 1 && size(v,2) != 1
-        throw(DimensionMismatch("input should be nx1 or 1xn"))
-    end
-
-    n = length(v)
-    numnz = nnz(v)
-    colptr = Vector{Ti}(n+1)
-    rowval = Vector{Ti}(numnz)
-    nzval = Vector{Tv}(numnz)
-
-    if size(v,1) == 1
-        copy!(colptr, 1, v.colptr, 1, n+1)
-        ptr = 1
-        for col = 1:n
-            if colptr[col] != colptr[col+1]
-                rowval[ptr] = col
-                nzval[ptr] = v.nzval[ptr]
-                ptr += 1
-            end
-        end
-    else
-        copy!(rowval, 1, v.rowval, 1, numnz)
-        copy!(nzval, 1, v.nzval, 1, numnz)
-        colptr[1] = 1
-        ptr = 1
-        col = 1
-        while col <= n && ptr <= numnz
-            while rowval[ptr] > col
-                colptr[col+1] = colptr[col]
-                col += 1
-            end
-            colptr[col+1] = colptr[col] + 1
-            ptr += 1
-            col += 1
-        end
-        if col <= n
-            colptr[(col+1):(n+1)] = colptr[col]
-        end
-    end
-
-    return SparseMatrixCSC(n, n, colptr, rowval, nzval)
-end
 
 # Sort all the indices in each column of a CSC sparse matrix
 # sortSparseMatrixCSC!(A, sortindices = :sortcols)        # Sort each column with sort()

base/sparse/sparsevector.jl

@@ -2004,3 +2004,30 @@ function fill!(A::Union{SparseVector, SparseMatrixCSC}, x)
     end
     return A
 end
+
+function diagm(v::SparseVector{Tv,Ti}) where {Tv,Ti}
+    n = length(v)
+    numnz = nnz(v)
+    colptr = Vector{Ti}(n+1)
+    rowval = Vector{Ti}(numnz)
+    nzval = Vector{Tv}(numnz)
+
+    copy!(rowval, 1, v.nzind, 1, numnz)
+    copy!(nzval, 1, v.nzval, 1, numnz)
+    colptr[1] = 1
+    ptr = 1
+    col = 1
+    while col <= n && ptr <= numnz
+        while rowval[ptr] > col
+            colptr[col+1] = colptr[col]
+            col += 1
+        end
+        colptr[col+1] = colptr[col] + 1
+        ptr += 1
+        col += 1
+    end
+    if col <= n
+        colptr[(col+1):(n+1)] = colptr[col]
+    end
+    return SparseMatrixCSC(n, n, colptr, rowval, nzval)
+end

test/sparse/sparse.jl

@@ -1316,13 +1316,6 @@ end
     @test trace(speye(5)) == 5
 end
 
-@testset "diagm on a matrix" begin
-    @test_throws DimensionMismatch diagm(sparse(ones(5,2)))
-    @test_throws DimensionMismatch diagm(sparse(ones(2,5)))
-    @test diagm(sparse(ones(1,5))) == speye(5)
-    @test diagm(sparse(ones(5,1))) == speye(5)
-end
-
 @testset "diag" begin
     for T in (Float64, Complex128)
         S1 = sprand(T,  5,  5, 0.5)

test/sparse/sparsevector.jl

@@ -1150,3 +1150,10 @@ end
 @testset "spzeros with index type" begin
     @test typeof(spzeros(Float32, Int16, 3)) == SparseVector{Float32,Int16}
 end
+
+@testset "diagm" begin
+    v = sprand(10, 0.4)
+    @test diagm(v)::SparseMatrixCSC                == diagm(Vector(v))
+    @test diagm(sparse(ones(5)))::SparseMatrixCSC  == speye(5)
+    @test diagm(sparse(zeros(5)))::SparseMatrixCSC == spzeros(5,5)
+end