Measures of Variation
方差:measures of variation or measures of spread
From range discovery range is not sufficient to assess the whole the SET (because only use Largest and Smallest value ), so with variance
The Sample Standard Deviation:
偏差:How far, on average, the observations are from the mean;
Geometric meaning:
Must take the absolute value, or else the second column SUM = 0 (because the average is obtained to find that the resulting count value and the sum of all the values of 0 point), so here taking the square, so the sum of squared deviations, the sum of squares .
Why sample variance divided by ( the n--1 )?
Other ways:
Rounding Rule: Do not perform any rounding until the computation is complete; otherwise, substantial roundoff error can result
With a standard deviation variance restored as a mean gap value, we proved this point (the first intuitively the Expect the Range )