Re: [ATM] FigureXP - monte carlo analysis

["Bob May" <rmay@nethere.com>]

One of the things that I've learned from doing data analysis in
electronics is that the results get strange and non-responsive to
what is actually happening when you don't get your data right.
You don't have to start at 0 if you really don't want to, but you
do have to make all of the runs match.  The easiest way is to set
the first value to 0 and that problem goes away in a spectaular
fashion.  I don't do the math but rather let one of the software
types do that stuff but I do run the programs and when you don't
normalize the data, nothing is right.  I'll even note that you
have to normalize the data correctly or you will also get
strangeness.  One job had a program that measured two signals and
measured range and range rate from that.  Early on, the program
kept on finding a 0 range target on some units.  Found out that
you have to do things in a particular order in the program to
prequalify the inputs or you'll get wrong answers.  Basic problem
was that software people don't understand analog stuff and the
hardware engineers didn't understand the things that the software
people were doing to the data.
Bob May

rmay at nethere.com
http: slash /nav.to slash bobmay
http: slash /bobmay dot astronomy.net

----- Original Message -----
From: Mitchell R <funnybone101@embarqmail.com>
To: <atm@atmlist.net>
Sent: Saturday, May 17, 2008 2:36 PM
Subject: Re: [ATM] FigureXP - monte carlo analysis


>
>
> -----Original Message-----
> From: atm-bounces@atmlist.net [mailto:atm-bounces@atmlist.net]
On Behalf Of
> Bob May
> Sent: Saturday, May 17, 2008 5:26 PM
> To: atm@atmlist.net
> Subject: Re: [ATM] FigureXP - monte carlo analysis
>
> Sorry, but I'd assumed that you'd be doing all of the reading
> sets with the center zone at the "0" spot on the dial.  Those
> Dial Indicators do allow for the scale to be turned by turning
> the bezel of the dial.  I always find the 0 and then set the
dial
> so that it reads 0.
> When you do put in random starts to a Monte Carlo routine, you
do
> indeed get a lot of strangeness.
> Bob May
>
> rmay at nethere.com
> http: slash /nav.to slash bobmay
> http: slash /bobmay dot astronomy.net
>
>
>      Why does everyone want to start at 0? It doesn't matter.
If you put in
> a data set .010, .020, .030, it is exactly that same as .110,
.120, .130, or
> .510, .520, .530. The deltas are important. Am I missing
something?
>
> Mitchell
>
> _______________________________________________
> ATM mailing list http://www.atmlist.net/
>

_______________________________________________
ATM mailing list http://www.atmlist.net/