Skip to main content

FLOAT_SCALE_EXP

mz_compute::render::reduce

Constant FLOAT_SCALE_EXP 

Source 
const FLOAT_SCALE_EXP: u32 = 24;
Expand description

The number of fractional bits of binary precision retained by the fixed-point representation used to accumulate float sums. The fixed-point scale is FLOAT_SCALE == 2^FLOAT_SCALE_EXP.