On Wed, 13 Aug 2003, Top Spin wrote in part:

> On 13 Aug 2003, [EMAIL PROTECTED] (Donald Burrill) wrote:
>
> >Why re-invent the wheel, and have to de-bug it besides?  Get one of
> >the standard statistical analysis packages (SPSS, SAS, MINITAB, ...)
> >and use the multiple regression routine to fit your data to a
> >suitable model.
>
> I certainly do not want to reinvent the wheel. I will eventually need
> custom code to run with my data live, ...

"Running live" could mean operating continuously;  but the nature of
statistical analysis rather requires a batch of data to work on.  One
could be collecting data on a quasi-continuous basis, and every so often
(after 15 minutes, say, or after 25 new data points) invoke something
statistical to see whether your process is still following the same
decay model.  One might also have something in place, based on the
current statistical model, to examine each new datum to see whether it
appears to depart worrisomely from the current model -- control chart
technology, essentially.

> but your suggestion of using a
> commercial package to analyze existing data is a good one.

The folks at Minitab Inc. (or SPSS, or ...) could probably tell you how,
and/or help you, to access the routine(s) you need as (a)
subroutine(s) to your data-gathering software.
 Or vice versa, should that be preferable for some reason.

> I have heard that Excel can do some statistical analyses. The help
> file says I need to install the Analysis Toolpak. Do you know how this
> compares with SPSS or Minitab? Can you recommend a good book on using
> Excel for this?

Hmph.  I cannot recommend Excel for this.  Haven't worked with it
myself, but am told that the statistical routines in the Excel add-on(s)
are unsatisfactory in several respects, including (I believe) precision
of results and stability of algorithms.  But I'll let colleagues on this
list who have more detailed information speak to that.

Personally, if I were using Excel as a convenient means of gathering and
storing data, I'd output the data in a plain text file, then read that
file in MINITAB to carry out any statistical analyses;  then extract
whatever I needed from the MINITAB output to adjust operating parameters
(or whatever) in the data-gathering mechanisms;  and so on.

< snip, the rest >

Good luck!   -- DFB.
 -----------------------------------------------------------------------
 Donald F. Burrill                                         [EMAIL PROTECTED]
 56 Sebbins Pond Drive, Bedford, NH 03110                 (603) 626-0816

.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to