Added pow function fpm #1621
Added pow function fpm #1621
Conversation
Implemented a potential working version of pow using fixed point math. No way to test because project won't run.
There was a problem hiding this comment.
Hey thanks for the contribution, but your code still contains floating point math which is not what #1543 was about. What we want are fuctions that implement these operations by only using fixed-point arithmetic.
I don't know if you want to give this another try but we can't really use this code right now.
| if (base == 0.0 && exponent == 0.0) { | ||
| return openage::util::FixedPoint<I, F>::quiet_NaN(); // Undefined; return NaN | ||
| } else if (base == 1.0 || exponent == 0.0) { |
There was a problem hiding this comment.
why compare to doubles in the if conditions when we could compre to fixed point values instead?
| double log_base = std::log(base.get_raw_value()); | ||
| double result = exponent * log_base; |
There was a problem hiding this comment.
The pure fixed-point math functions should avoid any floating point arithmetic. Avoiding the floating point rounding erros is the point of using fp math after all.
There was a problem hiding this comment.
this file would need to be removed
|
I'm closing this because this is not what we are looking for. |
Implemented a potential working version of pow using fixed point math. No way to test because project won't run.