Vq = interpn(X1, X2, ..., Xn, V, Xq1, Xq2, ..., Xqn)
Vq = interpn(V, Xq1, Xq2, ..., Xqn)
Vq = interpn(V)
Vq = interpn(V, k)
Vq = interpn(..., method)
Vq = interpn(..., method, extrapval)
| Parameter | Description |
|---|---|
| X1, ..., Xn | Sample grid points: vectors or ndgrid arrays. |
| V | Sample values: real or complex array. Extra dimensions represent additional data sets on the same grid. |
| Xq1, ..., Xqn | Query points. |
| method | 'linear', 'nearest', 'pchip', 'cubic', 'makima', or 'spline'. PCHIP is only for 1-D. |
| Parameter | Description |
|---|---|
| Vq | Interpolated values. |
interpn interpolates N-D gridded data using ndgrid conventions. The default grid is 1:size(V,i) in each dimension.
The cubic-family methods use a native tensor-product four-point stencil, with linear fallback on dimensions that have fewer than four samples. PCHIP remains restricted to the 1-D interpn syntax.
interpn(V) and interpn(V,k) refine the default grid. Query arrays of the same size are treated as scattered points; mixed-orientation vectors define a full grid.
V = [1 2; 3 4];
Vq = interpn(V, 1.5, 1.5)x = [10 20];
y = [1 3];
[X,Y] = ndgrid(x, y);
V = [1 2; 3 4];
Vq = interpn(X, Y, V, 15, 2)V = [1 2; 3 4];
Vq = interpn(V, 1)V = reshape(1:16, [2 2 2 2]);
Vq = interpn(V, 1.5, 1.5, 1.5, 1.5)