ndarray: implement rdiv (#292)

* ndarray: implement rdiv

e.g.

```julia
1 ./ mx.NDArray(Float32[1 2; 3 4])
```

* typo

NEWS.md

@@ -3,7 +3,7 @@
 ## New API
 
 * `deepcopy` for NDArray (#273)
-
+* `scalar ./ NDArray` is available now. (#292)
 * `fill` and `fill!` for NDArray (#TBD)
   An API correspond to Python's `mx.nd.full()`
 

src/ndarray.jl

@@ -682,25 +682,42 @@ function div_from!(dst :: NDArray, arg :: Union{Real, NDArray})
   end
 end
 
+"""
+Elementwise division of NDArray
+"""
+div(x::NDArray, y::NDArray) = _div(x, y)
+div(x::NDArray, s::Real) = _div_scalar(x, scalar=s)
+div(s::Real, x::NDArray) = _rdiv_scalar(x, scalar=s)
+
 import Base: /
 """
     ./(arg0 :: NDArray, arg :: Union{Real, NDArray})
 
 Elementwise dividing an `NDArray` by a scalar or another `NDArray` of the same shape.
 """
-@compatdot function Base.broadcast(::typeof(/), arg0 :: NDArray, arg :: Union{Real, NDArray})
-  ret = copy(arg0, context(arg0))
-  div_from!(ret, arg)
+@compatdot function Base.broadcast(::typeof(/), arg0 :: NDArray,
+                                   arg :: Union{Real, NDArray})
+  div(arg0, arg)
+end
+
+@compatdot function Base.broadcast(::typeof(/), arg0 :: Real, arg :: NDArray)
+  div(arg0, arg)
 end
 
 """
     /(arg0 :: NDArray, arg :: Real)
 
-Divide an `NDArray` by a scalar. Matrix division (solving linear systems) is not implemented yet.
+Divide an `NDArray` by a scalar.
+Matrix division (solving linear systems) is not implemented yet.
 """
-function /(arg0 :: NDArray, arg :: Real)
-  arg0 ./ arg
-end
+/(arg0 :: NDArray, arg :: Real) = div(arg0, arg)
+
+"""
+    /(arg0 :: Real, arg :: NDArray)
+
+Elementwise divide a scalar by an `NDArray`.
+"""
+/(arg0 :: Real, arg :: NDArray) = div(arg0, arg)
 
 
 """

test/unittest/ndarray.jl

@@ -264,8 +264,47 @@ function test_div()
   t6, a6 = rand_tensors(Float16, dims)
   scalar_large = 1e4
   @test reldiff(t6 / scalar_large, copy(a6 ./ scalar_large)) < 1e-1
+
+  let x = mx.NDArray([1 2; 3 4])
+    @test eltype(x) == Int
+    @test copy(x / 2)   == [0 1; 1 2]
+    @test copy(x / 2.5) == [0 1; 1 2]
+    @test copy(x / 2.9) == [0 1; 1 2]
+  end
 end
 
+
+function test_rdiv()
+  info("NDarray::rdiv")
+
+  info("NDarray::rdiv::Inf16")
+  let x = 1 ./ mx.zeros(Float16, 4)
+    @test copy(x) == [Inf16, Inf16, Inf16, Inf16]
+  end
+
+  info("NDarray::rdiv::Inf32")
+  let x = 1 ./ mx.zeros(Float32, 4)
+    @test copy(x) == [Inf32, Inf32, Inf32, Inf32]
+  end
+
+  info("NDarray::rdiv::Inf64")
+  let x = 1 ./ mx.zeros(Float64, 4)
+    @test copy(x) == [Inf64, Inf64, Inf64, Inf64]
+  end
+
+  info("NDarray::rdiv::Int")
+  let x = 1 ./ mx.NDArray([1 2; 3 4])
+    @test copy(x) == [1 0; 0 0]
+  end
+
+  info("NDarray::rdiv::Float32")
+  let x = 1 ./ mx.NDArray(Float32[1 2; 3 4])
+    y = 1 ./ Float32[1 2; 3 4]
+    @test reldiff(copy(x), y) < 1e8
+  end
+end  # function test_rdiv
+
+
 function test_gd()
   dims   = rand_dims()
   tw, aw = rand_tensors(dims)
@@ -551,6 +590,7 @@ end
   test_minus()
   test_mul()
   test_div()
+  test_rdiv()
   test_gd()
   test_saveload()
   test_clip()