4.2 Struct Type and Accessors

A guard ensures that the values field has length (* components width height), and that each size field is a nonnegative fixnum.

The default flomap constructor is perhaps the hardest to use. Instead, to construct a flomap from scratch, you should generally use make-flomap, make-flomap*, build-flomap or draw-flomap.

If x or y is out of bounds, this function returns 0.0. If k is out of bounds, it raises an error. The Conceptual Model section explains why k is treated differently.

If x or y is out of bounds, this function returns an flvector filled with 0.0. It always returns an flvector of length (flomap-components fm).

Like all other flomap functions that operate on real-valued coordinates, flomap-bilinear-ref regards known values as being at half-integer coordinates. Mathematically, if x = (+ i 0.5) and y = (+ j 0.5) for any integers i and j, then (flomap-bilinear-ref fm k x y) = (flomap-ref fm k i j).

Suppose fm is size w×h. If x ≤ -0.5 or x ≥ (+ w 0.5), this function returns 0.0; similarly for y and h. If k is out of bounds, it raises an error. The Conceptual Model section explains why k is treated differently.

The values returned by flomap-nonzero-rect can be sent to subflomap to trim away zero values. But see flomap-trim, which is faster for alpha-multiplied flomaps.

This function is used by some library functions, such as flomap-bilinear-ref, to index into already-destructured flomaps. From untyped code, applying this function is likely no faster than applying flomap-ref, because of extra contract checks.