Let viz(StepRange, vector) work too like plot(...) does. (#1688)

* Test/demo file for the whittaker smoothing function.

* Let viz(StepRange, vec) work too.

Author: joa-quim

src/imshow.jl

@@ -65,10 +65,12 @@ function imshow(arg1, x::AbstractVector{Float64}=Float64[], y::AbstractVector{Fl
 	elseif (isa(arg1, Array{UInt8}) || isa(arg1, Array{UInt16,3}))
 		Gi = mat2img(arg1; kw...)
 		call_img = true
-	elseif (isa(arg1, GDtype) || isa(arg1, GMTfv) || isa(arg1, Vector{GMTfv}) || (isa(arg1, VecOrMat{<:Real}) && size(arg1,2) <= 4) || (isa(arg1, Gdal.AbstractDataset) || isa(arg1, Gdal.AbstractGeometry)))
+	elseif (isa(arg1, GDtype) || isa(arg1, GMTfv) || isa(arg1, Vector{GMTfv}) || (isa(arg1, AbstractVecOrMat{<:Real}) && size(arg1,2) <= 4) || (isa(arg1, Gdal.AbstractDataset) || isa(arg1, Gdal.AbstractGeometry)))
 		(isa(arg1, Gdal.AbstractDataset) || isa(arg1, Gdal.AbstractGeometry)) && (arg1 = gd2gmt(arg1))
 		isa(arg1, Matrix{<:Real}) && (arg1 = size(arg1,2) == 1 ? cat_1_arg(arg1, true) : mat2ds(arg1))
-		isa(arg1, Vector{<:Real}) && (arg1 = mat2ds([1:length(arg1) arg1]))	# Make it a 2D dataset with the first col = 1:ncol(arg1))
+		if (isa(arg1, AbstractVector{<:Real}))
+			arg1 = isempty(x) ? mat2ds([1:length(arg1) arg1]) : mat2ds([arg1 x])
+		end
 		if isa(arg1, Vector{GMTfv})
 			ginfo = getbbox(arg1);
 			isFV3D = (size(arg1[1].verts, 2) > 2)

src/utils.jl

@@ -842,7 +842,7 @@ end
 Returns the coefficients for a polynomial p(x) of degree `n` that is the least-squares best fit for the data in y.
 The coefficients in p are in ascending powers, and the length of p is n+1.
 
-The `xscale` parameter is useful when needing to get coeeficients in different x units. For example when converting
+The `xscale` parameter is useful when needing to get coeficients in different x units. For example when converting
 months or seconds into years.
 """
 polyfit(D::GMTdataset, n::Int=size(x,1)-1; xscale=1) = polyfit(view(D.data, :,1), view(D.data, :,2), n, xscale=xscale)