Re: Real-time Calculation of Standard Deviation
- From: "Michael Harrington" <mikharr@xxxxxxxxxxx>
- Date: Sun, 22 Jan 2006 20:40:38 +1100
"gd1956" <gdalby1956@xxxxxxxxxxxxx> wrote in message
news:1137916269.535456.162060@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>A while back, I was working on a project where data (a wire diameter)
> was measured and displayed. I wanted to display the mean and standard
> deviation from the start of the measurement but couldn't store all the
> data values until the end of the measurement due to memory limitations.
> I worked out a formula for calculating the current standard deviation
> based on the new data value, the previous mean and the previous
> standard deviation.
> Can anyone tell me whether this is 'common knowledge' (I am not a
> mathematician) as I haven't seen it mentioned anywhere.
>
See
http://en.wikipedia.org/wiki/Standard_deviation#Rapid_calculation_methods
Especially take note of the formula for running sums!
You need only to keep in memory....
(i) previous sample size.
(ii) previous SumOf ( diameters^2)
(iii) previous ( SumOf (diameters))^2
hth Mick
.
- Follow-Ups:
- Re: Real-time Calculation of Standard Deviation
- From: gd1956
- Re: Real-time Calculation of Standard Deviation
- References:
- Real-time Calculation of Standard Deviation
- From: gd1956
- Real-time Calculation of Standard Deviation
- Prev by Date: Re: Cantorian pseudomathematics
- Next by Date: Re: skew symmetric map/comples geometry
- Previous by thread: Real-time Calculation of Standard Deviation
- Next by thread: Re: Real-time Calculation of Standard Deviation
- Index(es):
Relevant Pages
|