Fix bug in findmin! and findmax!

Also added tests/docs for those functions and exported them

Author: sglyon

base/exports.jl

@@ -521,6 +521,8 @@ export
     findin,
     findmax,
     findmin,
+    findmin!,
+    findmax!,
     findn,
     findnext,
     findprev,

base/reducedim.jl

@@ -325,12 +325,18 @@ function findminmax!{T,N}(f, Rval, Rind, A::AbstractArray{T,N})
     Rval, Rind
 end
 
-# findmin
+
+"""
+    findmin!(rval, rind, A, [init=true]) -> (minval, index)
+
+Find the minimum of `A` and the corresponding linear index along singleton
+dimensions of `rval` and `rind`, and store the results in `rval` and `rind`.
+"""
 function findmin!{R}(rval::AbstractArray{R},
                      rind::AbstractArray,
                      A::AbstractArray;
                      init::Bool=true)
-    findminmax!(LessFun(), initarray!(rval, typemax(R), init), rind, A)
+    findminmax!(LessFun(), initarray!(rval, MinFun(), init), rind, A)
 end
 
 function findmin{T}(A::AbstractArray{T}, region)
@@ -342,12 +348,17 @@ function findmin{T}(A::AbstractArray{T}, region)
             zeros(Int, reduced_dims0(A, region)), A)
 end
 
-# findmax
+"""
+    findmax!(rval, rind, A, [init=true]) -> (maxval, index)
+
+Find the maximum of `A` and the corresponding linear index along singleton
+dimensions of `rval` and `rind`, and store the results in `rval` and `rind`.
+"""
 function findmax!{R}(rval::AbstractArray{R},
                      rind::AbstractArray,
                      A::AbstractArray;
                      init::Bool=true)
-    findminmax!(MoreFun(), initarray!(rval, typemin(R), init), rind, A)
+    findminmax!(MoreFun(), initarray!(rval, MaxFun(), init), rind, A)
 end
 
 function findmax{T}(A::AbstractArray{T}, region)

doc/stdlib/collections.rst

@@ -347,6 +347,18 @@ Iterable Collections
 
    For an array input, returns the value and index of the minimum over the given dimensions.
 
+.. function:: findmax!(rval, rind, A, [init=true]) -> (maxval, index)
+
+   .. Docstring generated from Julia source
+
+   Find the maximum of ``A`` and the corresponding linear index along singleton dimensions of ``rval`` and ``rind``\ , and store the results in ``rval`` and ``rind``\ .
+
+.. function:: findmin!(rval, rind, A, [init=true]) -> (minval, index)
+
+   .. Docstring generated from Julia source
+
+   Find the minimum of ``A`` and the corresponding linear index along singleton dimensions of ``rval`` and ``rind``\ , and store the results in ``rval`` and ``rind``\ .
+
 .. function:: maxabs(itr)
 
    .. Docstring generated from Julia source

test/reducedim.jl

@@ -113,15 +113,23 @@ R = reducedim((a,b) -> a+b, [1 2; 3 4], 2, 0.0)
 rt = Base.return_types(reducedim, Tuple{Function, Array{Float64, 3}, Int, Float64})
 @test length(rt) == 1 && rt[1] == Array{Float64, 3}
 
+
 ## findmin/findmax
 A = [1.0 3.0 6.0;
      5.0 2.0 4.0]
-@test findmin(A, (1,)) == ([1.0 2.0 4.0], [1 4 6])
-@test findmin(A, (2,)) == (reshape([1.0,2.0], 2, 1), reshape([1,4], 2, 1))
-@test findmin(A, (1,2)) == (fill(1.0,1,1),fill(1,1,1))
-@test findmax(A, (1,)) == ([5.0 3.0 6.0], [2 3 5])
-@test findmax(A, (2,)) == (reshape([6.0,5.0], 2, 1), reshape([5,2], 2, 1))
-@test findmax(A, (1,2)) == (fill(6.0,1,1),fill(5,1,1))
+for (tup, rval, rind) in [((1,), [1.0 2.0 4.0], [1 4 6]),
+                          ((2,), reshape([1.0,2.0], 2, 1), reshape([1,4], 2, 1)),
+                          ((1,2), fill(1.0,1,1),fill(1,1,1))]
+    @test findmin(A, tup) == (rval, rind)
+    @test findmin!(similar(rval), similar(rind), A) == (rval, rind)
+end
+
+for (tup, rval, rind) in [((1,), [5.0 3.0 6.0], [2 3 5]),
+                          ((2,), reshape([6.0,5.0], 2, 1), reshape([5,2], 2, 1)),
+                          ((1,2), fill(6.0,1,1),fill(5,1,1))]
+    @test findmax(A, tup) == (rval, rind)
+    @test findmax!(similar(rval), similar(rind), A) == (rval, rind)
+end
 
 # issue #6672
 @test sum(Real[1 2 3; 4 5.3 7.1], 2) == reshape([6, 16.4], 2, 1)