% call flexinterpn_method with single data point
% the kernel (second output) is then a 1-by-2 cell array suitable for
% argument expansion in calls to flexinterpn
[null, kernel] = flexinterpn_method(1, 1, 'cubic');

% visualize kernel
plot(k{1});

% call flexinterpn with that kernel
r = rand(10, 10);
ri = flexinterpn(r, [Inf, Inf; 1, 1; 1, 1; 10, 10], k{:}, 0);