apfloat bias and unbias function should be symetric
- Dominant language
- C++
- Stars
- 1.9k
- Forks
- 283
- Avg merge
- 2d 10h
- Merged PRs (30d)
- 135
Description
Currently we have a variety of function for dealing w/ bias which takes a variety of argument types:
- https://google.github.io/xls/floating_point/#apfloatbias: apply bias to the given unbiased exponent
- takes just the unbiases exponent
- https://google.github.io/xls/floating_point/#apfloatunbiased_exponent: return the unbiased exponent from a given apfloat
- takes a whole apfloat value (even if the rest of the apfloat fields are not used)
- https://google.github.io/xls/floating_point/#apfloatexponent_bias: return the bias for the given apfloat type
- take just an apfloat sentinel for type inference
We should unify the different variants and have symetric pair for function (and methods) that can operate on different type of arguments.
Contributor guide
Assessment
This issue has not been assessed yet.