As the unofficial chairman of the National Riffle Association, I was seriously ticked off this morning to discover a major flaw in the LV implementation of the good, old Fisher Yates shuffle. Fortunately, when it matters to me, I never use code that I can not trace back to the source, so I have always used my own implementations. But myself, and one or perhaps two others, have tried to raise awareness of this potentially nifty VI. I have included a VI which repeatedly Riffles an array of integers 0..9. It then calculates the distribution of final position for each value as well as the distribution of values for each index in the riffled array. This should be flat, but notice the bias near the edges. For reference I have added a simple implementation of the FY shuffle, notice the lack of bias in the results. Other than license protection (Riffle is FDS+), why is the implementation of such a standard algorithm hiding in a CLFN? Why is the algorithm, like many others not referenced? Tested in LV9, LV10 on Windoze.

Darin.K, Great job and good find. Thanks to your efforts this was reported to R&D (# 344833) for further investigation. Thank you for the feedback! Andrew S.

LabVIEW

Riffle is Biased!

Solved!

Good to know its fixed. I reciently ran across a "Practical Exam" where riffle was advantagous. I felt bad about not implementing the workaround of ignoring index 0 But, didn't want to have to explain the W-A to the examiner.

"Should be" isn't "Is" -Jay

Message 21 of 21

(588 Views)

LabVIEW

Riffle is Biased!

Re: Riffle is Biased!