Remove workaround for pre-0.4.2 Julia support

Author: nalimilan

REQUIRE

@@ -1,4 +1,4 @@
-julia 0.4
+julia 0.4.2
 NamedArrays
 DataArrays
 DataFrames

src/freqtable.jl

@@ -145,75 +145,6 @@ function freqtable(x::PooledDataVector...; usena::Bool = false)
 	end
 end
 
-
-# Specialize one-way table until https://github.com/JuliaLang/julia/issues/13979
-# is backported to 0.4
-function freqtable(x::AbstractVector;
-                   # Parametric unions are currently not supported for keyword arguments,
-                   # so weights are restricted to Float64 for now
-                   # https://github.com/JuliaLang/julia/issues/3738
-                   weights::Union{Void, AbstractVector{Float64}} = nothing,
-                   subset::Union{Void, AbstractVector{Int}, AbstractVector{Bool}} = nothing)
-    if subset != nothing
-        x = y[subset]
-
-        if weights != nothing
-            weights = weights[subset]
-        end
-    end
-
-    l = (length(x),)
-
-    if weights != nothing && length(weights) != l[1]
-        error("'weights' (length $(length(weights))) must be of the same length as vectors (length $(l[1]))")
-    end
-
-    counttype = weights == nothing ? Int : eltype(weights)
-    d = Dict{eltype(x), counttype}()
-
-    for (i, el) in enumerate(x)
-        index = ht_keyindex(d, el)
-
-        if weights == nothing
-            if index > 0
-                d.vals[index] += 1
-            else
-                d[el] = 1
-            end
-        else
-            @inbounds w = weights[i]
-
-            if index > 0
-                d.vals[index] += w
-            else
-                d[el] = w
-            end
-        end
-    end
-
-    k = collect(keys(d))
-
-    s = Set{eltype(x)}()
-    for j in 1:length(k)
-        push!(s, k[j])
-    end
-
-    dimnames = unique(s)
-    T = eltype(dimnames)
-    if method_exists(isless, (T, T))
-        sort!(dimnames)
-    end
-
-    a = zeros(counttype, length(dimnames))
-    na = NamedArray(a, (dimnames,), ("Dim1",))
-
-    for (k, v) in d
-        na[k] = v
-    end
-
-    na
-end
-
 function freqtable(d::DataFrame, x::Symbol...; args...)
     a = freqtable([d[y] for y in x]...; args...)
     setdimnames!(a, x)