Raul Miller <[email protected]> writes: > J's standard deviation routine seems to already be dealing with the > numerical stability issue. > [...] > Or am I overlooking something?
J's standard deviation routine is perfectly accurate for any sane data. John Cook gave an (extreme) example where Welford's method gives a reasonable value the sum-of-squared-deviance method (the one used by J's stddev) breaks down: stddev 1e11+?1e6#0 0.538747 (He actually goes all the way to 1e12 rather than just 1e11.) Now, clearly, this is numerical extremism. I just wanted to reproduce the experiment myself. I still have to think my way through the examples that others have provided. Regards, Johann ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
