Archives of the TeradataForum

Message Posted: Fri, 30 Jan 2004 @ 21:10:44 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Rudel,

Not sure I am a math person but I tried to reproduce the formula and I make it as:

STDDEV_POP = SQRT{ [ COUNT(x)SUM(x**2) - (SUM(x))**2 ] / (COUNT(x))**2 }

I checked the formula with the simple population x1=0 and x2=2:

COUNT(x)=2
SUM(x)=2
SUM(x**2)=4

So, STDDEV_POP = SQRT{[2*4 - 2**2] / 2**2} = 1, which seems to be correct.

In this example I get SQRT(7) with both of the formulae you mention, so until somebody finds a counterexample, I will stick with my formula.

Regards,

Victor

	<Prev	Next>		<<First	<Prev	Next>	Last>>