From: Jo Butler <pjgb@mrc-lmb.cam.ac.uk>
  To  : arthur.rowe@nottingham.ac.uk
  Date: Wed, 02 Jun 1999 13:15:01 +0100

Re: radial step intervals in the XL scanning abs optics

Dear Arthur,

Your note stimulated me to try the same analysis on the radius values from
scans on 3 cells from a recent equilibrium run.  The nominal step size was
the default value of 0.003.

Analyses of the 3 data sets gave the following statistics:

=============================================
STATISTICS:  Cell 1

# Valid numbers       	=   	294
Sum                   	=   	0.9560
Mean                  	=   	3.3e-3
Variance              	=   	1.9e-6
Standard deviation    	=   	1.4e-3
Mean abs. deviation   	=   	1.1e-3
Skewness              	=   	0.5115
Kurtosis              	=   	0.2069
Median                	=   	3.0e-3
Maximum               	=   	8.0e-3
Minimum               	=   	1.0e-3
=============================================
STATISTICS:  Cell 2

# Valid numbers       	=   	294
Sum                   	=   	0.9500
Mean                  	=   	3.2e-3
Variance              	=   	2.1e-6
Standard deviation    	=   	1.5e-3
Mean abs. deviation   	=   	1.2e-3
Skewness              	=   	0.7008
Kurtosis              	=   	0.8777
Median                	=   	3.0e-3
Maximum               	=   	9.0e-3
Minimum               	=   	1.0e-3
=============================================
STATISTICS: Cell 3

# Valid numbers       	=   	286
Sum                   	=   	0.9000
Mean                  	=   	3.1e-3
Variance              	=   	1.8e-6
Standard deviation    	=   	1.3e-3
Mean abs. deviation   	=   	1.0e-3
Skewness              	=   	0.7217
Kurtosis              	=   	0.8707
Median                	=   	3.0e-3
Maximum               	=   	8.0e-3
Minimum               	=   	1.0e-3

Like yours, our XL-A is regularly serviced and I had done a radial
calibration on reaching speed at the start of the run.
The actual means are rather closer to the set value (3 um) than you found,
i.e. 3.3, 3.2 and 3.1 um, but the minimum is 1um in each case and the
maxima are 8 or 9 um and the standard deviations are ~50% of the set value.
Again the surprising point that the median is the statistic closest to the
set value.

Jo


--On Wed, Jun 2, 1999 12:23 pm +0000 Arthur Rowe
<Arthur.Rowe@nottingham.ac.uk> wrote:

> Hi RASMBers !
> 
> Trying to do a bit of hard-nosed analysis of data from scanning
> absorption optics,  and wanting to have data equi-spaced in r (well -
> thats nice for all sorts of good,  sound, numerico-analytical reasons) I
> was struck by how variable the step size in r  seemed to be within a
> given run.
> 
> With a nominal set interval of 2 micrometres, for a data set quite typica
> of many, I  found the following:
> 
> =============================================
> STATISTICS FOR FIRST DIFFERENCE IN r:
> 
> # Valid numbers       	=   	53
> Sum                   	=   	0.1440
> Mean                  	=   	2.7170e-3
> Variance              	=   	2.1684e-6
> Standard deviation    	=   	1.4725e-3
> Mean abs. deviation   	=   	1.2374e-3
> Skewness              	=   	0.5906
> Kurtosis              	=  	-0.8024
> Median                	=   	2.0003e-3
> Maximum               	=   	6.0000e-3
> Minimum               	=   	9.9993e-4
> =============================================
> 
> In other words, the ACTUAL mean value was 2.72 micrometres, with a
> standard  deviation of 1.47 micrometres, and individual values varying
> between 1 and 6  micrometres !!!  Oddly enough, the MEDIAN value is
> almost precisely 2  micrometres.
> 
>  I should say we have our XL-A regularly serviced, and the slit scanning
>  system was  specifically checked (and its radial calibration) not long
> ago.
> 
> What is the general experience here of XL users ?  Is this just one of
> those things  one has to put up with ?
> 
> Best wishes to all
> 
> Arthur Rowe
> 
> 
> 
> *************************************************
> Professor Arthur J Rowe
> NCMH Business Centre
> University of Nottingham
> School of Biological Sciences
> Sutton Bonington
> Leicestershire LE12 5RD   UK
> 
> phone/voicemail					+44 (0)115 951 6156
> phone/telefax							    +44 (0)115 951 6157
> email														            	arthur.rowe@nottingham.ac.uk
> www.nottingham.ac.uk/ncmh/business
> *************************************************



Dr P.J.G. Butler,
MRC Laboratory of Molecular Biology,
Hills Road,
Cambridge, CB2 2QH,
UK.
Tel. 01223 248011 (or 01223 402296 DDI)
FAX. 01223 213556