Thanks for your work, this is a very interesting project!
I'm not sure if my thinking is correct, so I'd like to ask you if it would make sense to limit the maximum order for continuous mappings.
For the f32 mapping, the macro defining that aspects expands into:
pub(crate) const ORDER_MAX: u32 = U32_BITS / 2;
In which case the ORDER_MAX would be 16.
But in reality the maximum integer precision for f32 is 16777216 = 4096² which would indicate an order of 12. After 4096² the f32 no longer can represent odd integers so I'm not sure if it would be a viable tradeoff, on continuing the hilbert approximation vs start losing integer-precision indexing.
Then again I'm not sure if I'm missing something, but I thought of making this suggestion nevertheless. Thanks once again!
Thanks for your work, this is a very interesting project!
I'm not sure if my thinking is correct, so I'd like to ask you if it would make sense to limit the maximum order for continuous mappings.
For the
f32
mapping, the macro defining that aspects expands into:In which case the
ORDER_MAX
would be 16.But in reality the maximum integer precision for f32 is 16777216 = 4096² which would indicate an order of 12. After 4096² the f32 no longer can represent odd integers so I'm not sure if it would be a viable tradeoff, on continuing the hilbert approximation vs start losing integer-precision indexing.
Then again I'm not sure if I'm missing something, but I thought of making this suggestion nevertheless. Thanks once again!