Re: Mean Vs. Geometric Mean calculation correction

From: Ray Hicks (rh208@cam.ac.uk)
Date: Mon May 07 2001 - 08:17:53 EST


>
> A term is missing here: the number of decades in the scale.  So, the
> equation should read:
> linear value =10^ (4*(channel value/256))

There isn't a missing term, and the equation shouldn't read that way, except
in the erroneous instance below (Re-read the paragraph following the
equation that you "corrected" above, and you'll notice that I mentioned that
256 should be replaced by 64 for 8 bit data on a 4 decade scale)


>
> Also note that the 256 is the channel resolution of the instrument (or if
> you are using WinList it is the histogram resolution). So it may be 32, 64,
> 128, 256 or 1024 depending on your application.
>

No it isn't - the FACScan is a ten bit instrument, and FACScan Research
software records ten bits (1024 channels), on a four decade scale. (The case
you put at the end)

> This stuff gives us fits too!

I'm sure it does!


>
> Albert Donnenberg
> University of Pittsburgh
>
Ray


>



This archive was generated by hypermail 2b29 : Sun Jan 05 2003 - 19:01:17 EST