On 02/07/2016 09:08, Alex Bennée wrote: > Hmm I was doing from memory but it should be the mean of the sum of the > squares of the deviation: > > # calculate deviation > deviation = 0 > for r in res: > deviation += (r.time - avg_time)**2 > > deviation = deviation / count
This is the population variance. "Standard deviation" and "variance" usually refer to the sample standard deviation and sample variance, where you divide by count-1. For the population stdev and sample stdev you have to take the square root. LibreOffice agrees with Emilio, giving a sample stdev of 0.2742 and a population stdev of 0.2453 Paolo
