incorrect constant in pcg_output_xsh_rr_128_64?

[simonbyrne]

Is the shift of 29 here:
https://github.com/imneme/pcg-c/blob/e2383c4bfcc862b40c3d85a43c9d495ff61186cb/include/pcg_variants.h#L166
correct?

From what I can tell from the paper, it should be floor((6+64)/2) = 35?

[sunoru]

With the algorithm here the shift is actually 35, rather than 29.

So the expected files of tests are also wrong.

[imneme]

I'll look into it, but it does seem like you've found an issue!

Thanks!

Melissa.

On Jun 14, 2016, at 7:04 AM, スノル [email protected] wrote:

With the algorithm here the shift is actually 35, rather than 29.

So the expected files of tests are also wrong.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or mute the thread.