Closed glennsweeney closed 10 years ago
These constants don't seem to have full precision. Is that intentional?
I chose the 7 digits of precision as that was what I found elsewhere in the file. However, after looking further, those 7 digits seem to be based off of published values, not internal computation.
In all honesty, I don't believe the additional precision will make a real-world difference, as the DIN standards only provide 5 significant figures for their constants. However, the most recent commit that I pushed updates them to match Julia's output for cosd()
and sind()
.
Also, thanks again @StefanKarpinski for your help and guidance.
Got it, @glenn-sweeney. Looks good.
Minor fixes to the conversion functions. I'm not positive if this will improve performance significantly or not with the Julia JIT, but it's generally good practice to hard-code constants anyhow. Also provided better commenting as to why
kch
andke
are set to 1; the standard provides for them, but does not recommend their use. If in the future someone really needs them, they are available.