Standard Deviation and Variance Weighting

[Jonnie_5]

Can anyone give a quick explanation of the weighting node on the standard dev/var vi? What's the difference between a 'sample' and a 'population' weighting? I figured if you toss the vi an array, it would always be a population.

I shouldn't have slept through stats...

[MartinManscher]

First, look at the diagram for the standard dev/var vi. The mean is the same regardless of the weighting. If I remember correctly, the difference is that "Sample" assumes that the data is taken out of a larger population of data. Having n-1 in the denominator makes the standard deviation result a maximum likelihood estimator for the real standard deviation (of the whole population). Technically speaking, you use n-1 instead of n because computing the average takes of one degree of freedom from the data set.

However, if you choose "population" you tell the VI that the data you give is the whole population, in which case you don't need an estimator, you can compute the real standard deviation.

Usually your data is not the whole population
, so you should use "sample" weighting.

[Jonnie_5]

Very cool.
!Muchas gracias!
~js