Skip to content

nls: singular convergeance

5 messages · Rolf Turner, Redding, Matthew

Redding, Matthew
# 
Hi Rolf and others

Will this make any difference to the convergence?

Thanks

Matt



Matt Redding, Ph.D.
Soil chemist/Geochemist
Principal Scientist
Agri-Science Queensland
Ph 0746 881372
Mob 0408 787100
Fax 07 46881192

----- Original Message -----
From: Rolf Turner <rolf.turner at xtra.co.nz>
To: Redding, Matthew
Cc: r-help at r-project.org <r-help at r-project.org>
Sent: Thu Oct 13 18:04:53 2011
Subject: Re: [R] nls: singular convergeance
On 13/10/11 18:02, Redding, Matthew wrote:
<SNIP>

Have you thought of using the "segmented" package instead
of rolling your own with nls?

     cheers,

         Rolf Turner

We're behind the Bid!
GOLD COAST 2018 - XXI COMMONWEALTH GAMES CANDIDATE CITY
www.goldcoast2018bid.com

********************************DISCLAIMER****************************
The information contained in the above e-mail message or messages 
(which includes any attachments) is confidential and may be legally 
privileged.  It is intended only for the use of the person or entity 
to which it is addressed.  If you are not the addressee any form of 
disclosure, copying, modification, distribution or any action taken 
or omitted in reliance on the information is unauthorised.  Opinions 
contained in the message(s) do not necessarily reflect the opinions 
of the Queensland Government and its authorities.  If you received 
this communication in error, please notify the sender immediately 
and delete it from your computer system network.
Rolf Turner
#
On 13/10/11 21:26, Redding, Matthew wrote:
I have no idea, really.  I don't know if convergence issues even
     arise in the context of the "segmented" package.  I have never
     encountered any.

     Try it and see.  Shouldn't take more than about 10 minutes, all up.

         cheers,

             Rolf
Redding, Matthew
# 
Hi Rolf and Others, 

Thanks Rolf for your suggestion.
I've had a look at segmented, but can't see how it can help.

My model is the solution to an ODE of a physical system. The fitted
parameters from 
this model are the answers I require. 

I'm not sure how segmented can help -- it is 
a linear regression tool. Can you let me know how it could be used?

Thanks, 

Matt Redding
We're behind the Bid!
GOLD COAST 2018 - XXI COMMONWEALTH GAMES CANDIDATE CITY
www.goldcoast2018bid.com

********************************DISCLAIMER****************************
The information contained in the above e-mail message or messages 
(which includes any attachments) is confidential and may be legally 
privileged.  It is intended only for the use of the person or entity 
to which it is addressed.  If you are not the addressee any form of 
disclosure, copying, modification, distribution or any action taken 
or omitted in reliance on the information is unauthorised.  Opinions 
contained in the message(s) do not necessarily reflect the opinions 
of the Queensland Government and its authorities.  If you received 
this communication in error, please notify the sender immediately 
and delete it from your computer system network.
Rolf Turner
#
On 14/10/11 11:39, Redding, Matthew wrote:
No I can't. You said you were trying to fit a "broken stick" model,
which is what the "segmented" package does.  If you are not
fitting a model to data, then I have no idea as to how you might
proceed.

You have to make it much more clear what you are trying to do
before I (or anyone) can make constructive suggestions.

     cheers,

         Rolf Turner
Redding, Matthew
# 
Hi Rolf and others, 

Once again, I think a misunderstanding. My inexperience has led me to
use the 
term "broken stick" where "non-linear segments" might be more
appropriate.

This is all experimental data (t and seg_an), and yes I am fitting a
model to this data (see the "crv" function in
my example).

Fitting three linear segments to the data won't provide the parameter
estimates required.
The functions suiting this data are theoretically non-linear.  However,
I guess I have learned 
that ultimately I needed 
to collect measurements for longer until the non-linearity of 
each segment becomes more obvious. 

So I guess no amount of tweaking nls will help!

Thanks for your time.

Kind regards, 

Matt Redding
We're behind the Bid!
GOLD COAST 2018 - XXI COMMONWEALTH GAMES CANDIDATE CITY
www.goldcoast2018bid.com

********************************DISCLAIMER****************************
The informa...{{dropped:12}}