Prev: using mask with boundary conditions
Next: GUIDE gui handles visibility to user defined function
From: Stefan Andreevski on 6 May 2010 19:12
"Laurentiu " <laurREMOVECAPSmarinovici(a)gmail.com> wrote in message <hrv9kp$cee$1(a)fred.mathworks.com>...
> Hey
> Does anybody have any suggestions about how I could let MATLAB know that it should consider a number like 5.67*10^(-5) equal to 0, especially since, for example, all the other coefficients of the polynomial are of at least 10^(-2) order?
>
> Thanks

Hey Laurentiu,

A really easy trick would be just to decide your tollerance and do some simple rounding.
Let's say you allow at most 3 consecotive zeros after the decimal point (e.g. 0.000123)
Then your tol will be 1000000. Multiply all your data by your tolerance
e.g.: 0.000123 x 1000000 = 123
and then round it, ceil it, floor it or whatever you would choose. Again divide it by your multiplier
A pole like -0.0000000123 will just become 0.123 after multiplication and then 0 after rounding. Divide zero by your multiplier and again you will get a zero.
You lose a lot of precision using this approach, but I am also doing control engineering and I know precission sometimes messes things more than it actually helps.

Cheers,
Stefan
From: Laurentiu on 6 May 2010 21:43
"Stefan Andreevski" <standreevski(a)gmail.com> wrote in message <hrvic4$d0s$1(a)fred.mathworks.com>...
> "Laurentiu " <laurREMOVECAPSmarinovici(a)gmail.com> wrote in message <hrv9kp$cee$1(a)fred.mathworks.com>...
> > Hey
> > Does anybody have any suggestions about how I could let MATLAB know that it should consider a number like 5.67*10^(-5) equal to 0, especially since, for example, all the other coefficients of the polynomial are of at least 10^(-2) order?
> >
> > Thanks
>
> Hey Laurentiu,
>
> A really easy trick would be just to decide your tollerance and do some simple rounding.
> Let's say you allow at most 3 consecotive zeros after the decimal point (e.g. 0.000123)
> Then your tol will be 1000000. Multiply all your data by your tolerance
> e.g.: 0.000123 x 1000000 = 123
> and then round it, ceil it, floor it or whatever you would choose. Again divide it by your multiplier
> A pole like -0.0000000123 will just become 0.123 after multiplication and then 0 after rounding. Divide zero by your multiplier and again you will get a zero.
> You lose a lot of precision using this approach, but I am also doing control engineering and I know precission sometimes messes things more than it actually helps.
>
> Cheers,
> Stefan

Stefan,
Thanks a lot for your suggestion. And,actually, thanks to all of you guys for all suggestions. I will take them all into consideration and hopefully get my stuff done.

Cheers
From: dpb on 7 May 2010 10:19
Laurentiu wrote:
> "Stefan Andreevski" <standreevski(a)gmail.com> wrote in message
> <hrvic4$d0s$1(a)fred.mathworks.com>...
>> "Laurentiu " <laurREMOVECAPSmarinovici(a)gmail.com> wrote in message
>> <hrv9kp$cee$1(a)fred.mathworks.com>...
>> > Hey
>> > Does anybody have any suggestions about how I could let MATLAB know
>> that it should consider a number like 5.67*10^(-5) equal to 0,
>> especially since, for example, all the other coefficients of the
>> polynomial are of at least 10^(-2) order?
>> > > Thanks
>>
>> Hey Laurentiu,
>>
>> A really easy trick would be just to decide your tollerance and do
>> some simple rounding.
>> Let's say you allow at most 3 consecotive zeros after the decimal
>> point (e.g. 0.000123)
>> Then your tol will be 1000000. Multiply all your data by your tolerance
>> e.g.: 0.000123 x 1000000 = 123 and then round it, ceil it, floor it or
>> whatever you would choose. Again divide it by your multiplier
>> A pole like -0.0000000123 will just become 0.123 after multiplication
>> and then 0 after rounding. Divide zero by your multiplier and again
>> you will get a zero.
>> You lose a lot of precision using this approach, but I am also doing
>> control engineering and I know precission sometimes messes things more
>> than it actually helps.
>>
>> Cheers,
>> Stefan
>
> Stefan,
> Thanks a lot for your suggestion. And,actually, thanks to all of you
> guys for all suggestions. I will take them all into consideration and
> hopefully get my stuff done.

I'd point out that if your case is always that of the intercept term
that estimating the coefficients w/o it will have the side benefit of
the other coefficients being estimated under that condition instead of
simply truncating one ex post facto....

--
First  |  Prev  | 
Pages: 1 2 3
Prev: using mask with boundary conditions
Next: GUIDE gui handles visibility to user defined function