I think this might be what you want. Kate Mullen and I have been in
correspondence over some edge cases where minpack.LM may not handle bounds
appropriately. However, though nlmrt seems to do the job here, readers
should note that R benefits hugely if we maintain some friendly
competition (and c
Apologies if this is posted twice. The r-help mailing system gave an error
(reported to moderator) on first try, but it may have gone through.
Original Message
Subject: Re: R-help Digest, Vol 95, Issue 17
From:"Prof. John C Nash"
Date:
Ben and Ravi have already pointed out that you have a problem with a
non-computable (e.g., divide by zero or similar) objective function. This
is common.
optim() has mostly unconstrained optimizers. L-BFGS-B does handle box or
bounds constraints. There is also Rvmmin, which is supposedly the same
For nls, the fixing (or masking) of parameters is not, to my knowledge,
possible.
This is something I've been trying to get in such routines for over 2
decades. Masks are
available, but not yet well documented, in Rcgmin and Rvmmin packages.
However, these use
an optim() style approach, which is q
This is not really an answer to the original post, but an alternative
possibility that might benefit the R community as well as offering some
commercial revenue to someone with entrepreneurial instincts.
We have several web-based and similar interfaces to R. It should be
possible to have a server
Patrick Burns likely is closest to the truth in noting that the editing of
the NYT article was possibly savage. The author is probably fuming, and
can't do much or he'll not get future work.
I was a columnist for Interface Age and then a sub-editor for Byte in the
early 80s. If an ad came in close
I won't burden the list with copies of earlier posts -- all of us have
experienced the frustration of dealing with folk who want to make their
life easier by making ours difficult. However, I have noted that a few
folk are starting to change attitudes. I was hired to give a training
session last Ju
The optim() function really does no "statistics", so outlier detection is
not part of it. It is a link to function minimization. Ravi Varadhan and I
are doing some work to see what might be feasible to do to improve optim()
and unify the many optimization tools to help users choose the tools more
e
dhabby wrote:
Last week I run in to a lot a problems triyng to fit an ARIMA model to a
time series. The problem is that the internal process of the arima
function
call function "optim" to estimate the model parameters, so far so good...
but my data presents a problem with the defaul
> I am trying to fit a exponential power distribution
>
> y = b/(2*pi*a^2*gamma(2/b))*exp(-(x/a)^b)
>
> to a bunch of data for x and y I have in a table.
>> data
>x y
> 1 2527
> 2 7559
> 3125 219
> ...
> 25912925 1
> 260
This comment is orthogonal to most of the others. It seems that folk often
want to test for equality of "real" numbers. One important one is for
convergence tests. When writing my Compact Numerical Methods book I had to
avoid lots of logical tests, but wanted to compare two REALs. I found that
the
11 matches
Mail list logo
