Move describe functions for DataArrays from DataFrames to DataArrays (#248)

Author: cjprybol

REQUIRE

@@ -1,5 +1,5 @@
 julia 0.6-
 Compat 0.8.6
-StatsBase 0.3
+StatsBase 0.14.1
 Reexport
 SpecialFunctions

src/statistics.jl

@@ -35,3 +35,31 @@ function gl(n::Integer, k::Integer, l::Integer)
 end
 
 gl(n::Integer, k::Integer) = gl(n, k, n*k)
+
+StatsBase.describe(X::DataVector) = StatsBase.describe(STDOUT, X)
+
+function StatsBase.describe{T<:Real}(io::IO, X::AbstractDataVector{T})
+    nacount = sum(isna.(X))
+    pna = 100nacount/length(X)
+    if pna != 100 # describe will fail if dropna returns an empty vector
+        describe(io, dropna(X))
+    else
+        println(io, "Summary Stats:")
+        println(io, "Type:           $(eltype(X))")
+    end
+    println(io, "Number Missing: $(nacount)")
+    @printf(io, "%% Missing:      %.6f\n", pna)
+    return
+end
+
+function StatsBase.describe(io::IO, X::AbstractDataVector)
+    nacount = sum(isna.(X))
+    pna = 100nacount/length(X)
+    println(io, "Summary Stats:")
+    println(io, "Length:         $(length(X))")
+    println(io, "Type:           $(eltype(X))")
+    println(io, "Number Unique:  $(length(unique(X)))")
+    println(io, "Number Missing: $(nacount)")
+    @printf(io, "%% Missing:      %.6f\n", pna)
+    return
+end

test/statistics.jl

@@ -1,3 +1,69 @@
 @testset "Stats" begin
     autocor(DataArray([1, 2, 3, 4, 5]))
 end
+
+@testset "describe" begin
+    io = IOBuffer()
+    describe(io, @data(collect(1:10)))
+    @test String(take!(io)) == """
+                               Summary Stats:
+                               Mean:           5.500000
+                               Minimum:        1.000000
+                               1st Quartile:   3.250000
+                               Median:         5.500000
+                               3rd Quartile:   7.750000
+                               Maximum:        10.000000
+                               Length:         10
+                               Type:           $Int
+                               Number Missing: 0
+                               % Missing:      0.000000
+                               """
+    describe(io, @data([1, NA]))
+    @test String(take!(io)) == """
+                               Summary Stats:
+                               Mean:           1.000000
+                               Minimum:        1.000000
+                               1st Quartile:   1.000000
+                               Median:         1.000000
+                               3rd Quartile:   1.000000
+                               Maximum:        1.000000
+                               Length:         1
+                               Type:           $Int
+                               Number Missing: 1
+                               % Missing:      50.000000
+                               """
+    describe(io, @data(["s"]))
+    @test String(take!(io)) == """
+                               Summary Stats:
+                               Length:         1
+                               Type:           String
+                               Number Unique:  1
+                               Number Missing: 0
+                               % Missing:      0.000000
+                               """
+    describe(io, @data(["s", NA]))
+    @test String(take!(io)) == """
+                               Summary Stats:
+                               Length:         2
+                               Type:           String
+                               Number Unique:  2
+                               Number Missing: 1
+                               % Missing:      50.000000
+                               """
+    describe(io, DataArray(Any, 5))
+    @test String(take!(io)) == """
+                               Summary Stats:
+                               Length:         5
+                               Type:           Any
+                               Number Unique:  1
+                               Number Missing: 5
+                               % Missing:      100.000000
+                               """
+    describe(io, DataArray(Float64, 5))
+    @test String(take!(io)) == """
+                               Summary Stats:
+                               Type:           Float64
+                               Number Missing: 5
+                               % Missing:      100.000000
+                               """
+end