powered by
read float from binary (single-precision, 32 bits)
read_float(raw, cursor, n = 1L)
raw vector
offset
number of elements