forts. I look forward to the results.
Best regards,
Michael
unaffiliatd
> Gesendet: Freitag, 21. Oktober 2016 um 09:39 Uhr
> Von: "Berend Hasselman"
> An: "Mike meyer" <1101...@gmx.net>
> Cc: ProfJCNash , "r-help@r-project.org"
>
> Betr
Berend's point is well-taken. It's a lot of work to re-jig a code, especially
one more than
30 years old.
On the other hand, nlmrt is all-R, and it does more or less work on
underdetermined systems as I
illustrated in a small script. The changes needed to treat the problem as Mike
suggests are
> On 21 Oct 2016, at 06:00, Mike meyer <1101...@gmx.net> wrote:
>
> Let's take a different view of the problem.
> Given f=(f_1,...,f_m):R^n -> R^m we want to minimize ||f(x)||.
>
> What distinguishes this from a general minimization problem is that you know
> the structure of the
> objective fu
Cc: "r-help@r-project.org" , "Berend Hasselman"
>
> Betreff: Re: [R] nls.lm
>
> From a statistician's point of view, "nonsense" may be OK, but there are
> other applications of R where
> (partial or non-unique) solutions may be needed.
>
&
>From a statistician's point of view, "nonsense" may be OK, but there are other
>applications of R where
(partial or non-unique) solutions may be needed.
Yesterday I raised the question of how nonlinear least squares could be adapted
to underdetermined problems.
Many folk are unaware of such pos
> How do you reply to a specific post on this board instead of the thread?
You can reply to the individual, as I just did.
But I strongly suggest that you don't. You would be much better advised to
discontinue debate and follow the essential advice given by nls.lm, which - no
matter whether cou
> On Oct 19, 2016, at 12:16 PM, Mike meyer <1101...@gmx.net> wrote:
>
> How do you reply to a specific post on this board instead of the thread?
> I am too incompetent to find this out myself.
Most of us use a mail-client that supports 'reply-to-all'. And the Posting
Guide asks you to include s
How do you reply to a specific post on this board instead of the thread?
I am too incompetent to find this out myself.
Thanks,
Michael
unaffiliated
__
R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listin
> On 19 Oct 2016, at 19:12, Mike meyer <1101...@gmx.net> wrote:
>
>> From my reading of the above cited document I get the impression that the
>> algorithm
> (algorithm 3.16, p27) can easily be adapted to handle the case m In this case the Jacobian Jf(x) is mxn and the matrix A(x)=Jf(x)'Jf(x) is
And finally, to put to rest the notion that the number of residuals is in any
way significant for the
solution of the least squares problem I submit to you the function
f(x,y)=(x²+y²)²
of 2 variables but only one residual f_1(x,y)=x²+y² which nonetheless has a
unique minimum
at the point (0,0).
> On 19 Oct 2016, at 17:47 , Mike meyer <1101...@gmx.net> wrote:
> Jf(x)'Jf(x) nonsingular, for all x, is a reasonable condition, m>=n is not.
If Jf(x) has more columns than rows, then Jf(x)'Jf(x) is certainly singular.
The reverse is not true, but what's wrong with a simple pre-check?
What yo
>From my reading of the above cited document I get the impression that the
>algorithm
(algorithm 3.16, p27) can easily be adapted to handle the case m 0 and so the system becomes ill conditioned.
Why can we not get around this as follows: as soon as mu is below some threshold
we solve instead the
I sometimes find it useful to use nonlinear least squares for fitting an
approximation i.e., zero
residual model. That could be underdetermined.
Does adding the set of residuals that is the parameters force a minimum length
solution? If the
equations are inconsistent, then the residuals apart fr
Make that f(x,u)=||x||².
__
R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-c
@SE: yes, not every system of equations with more variables than equations is
solvable,
we need an additional condition e.g. full rank of the coefficient matrix.
Uniqueness of the solution was not required.
@BH:
Yes this is the document, it is a nice presentation.
I did not read the first page b
> On 19 Oct 2016, at 14:09, Mike meyer <1101...@gmx.net> wrote:
>
> @pd: you know that a System of equations with more variables than equations
> is always solvable
> and if a unique solution is desired one of mimimal norm can be used.
>
Not true.
Take the system with 3 variables and 2 equati
> Mike meyer
> @pd: you know that a System of equations with more variables than equations
> is always solvable
> and if a unique solution is desired one of mimimal norm can be used.
Here's an example of what you _said_:
x + y + z = 2
3x - y +4z = 4
Find a unique single-valued solution for al
> On 19 Oct 2016, at 14:09, Mike meyer <1101...@gmx.net> wrote:
>
> @pd: you know that a System of equations with more variables than equations
> is always solvable
> and if a unique solution is desired one of mimimal norm can be used.
>
You won't get a minimum norm solution by using a least
Peter is right that the conditions may be embedded in the underlying code. (Ask
Kate!)
My nlmrt package is all in R, so the conditions are visible. I'm currently in
process of rejigging this
using some work Duncan Murdoch helped with a while ago (I've had some other
things get in the way), so
I
@pd: you know that a System of equations with more variables than equations is
always solvable
and if a unique solution is desired one of mimimal norm can be used.
According to "Methods for nonlinear least squares problems" by Madsen, Nielsen
and Tingleff the LM-algorithm
solves Systems of the f
This would seem to apply to the add-on package minpack.lm. That package has a
maintainer...
Offhand, I would expect that this is a sanity check that, broadly speaking,
prevents you from trying to solve a system of equations with more unknowns than
equations. This is not a sufficient condition:
Greetings,
The description of nls.lm specifies that in minimizing a sum of squares of
residuals
the number of residuals must be no less than the dimension of the independent
variable
("par").
In fact the algorithm does not work otherwise (termination code 0).
But this condition is sensel
I will consider putting methods for AIC and logLik into the next version
of minpack.lm (contributions welcome).
For now, the following should work for logLik, where 'object' is the
return value of nls.lm.
logLik.nls.lm <- function(object, REML = FALSE, ...)
{
res <- object$fvec
N <- lengt
Hi there,
I'm a PhD student investigating growth patterns in fish. I've been using the
minpack.lm package to fit extended von Bertalanffy growth models that include
explanatory covariates (temperature and density). I found the nls.lm comand a
powerful tool to fit models with a lot of parameters
24 matches
Mail list logo
