The two packages have different purposes but for this specific gridding problem, GMT surface solution (using splines under tension) would provide a deterministic solution while the GeoStats package would provide a probabilistic one (by exploiting any spatial correlations present by using krigging for instance).
See an interesting discussion on this topic here.
Thanks for the summary.
I’ve tried the code as being described here to plot some Topographic EEG/MEG plot using Julia. The results are not quite satisfying, however.
The methods are described as follows in MatLab: griddata .
I’m wondering whether it results from the method difference to interpolate values?
@likanzhan, you may check other options in this other thread, in particular GMT.jl and DIVAnd.jl.
NB:
GMT’s splines in tension method also produces “surfaces that always pass through the data points”