Does it make any statistical sense to do polr or probit regression (not
the same thing) on `really huge data'? There are few regression-like
problems in which model inadequacy does not swamp estimation uncertainty
for as few as a 1000 cases.
If you want to do that sort of thing, by all means
Earl F. Glynn asks:
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Earl F. Glynn
Sent: Tuesday, 15 August 2006 8:44 AM
To: r-help@stat.math.ethz.ch
Subject: [R] Help with workaround for: Function '`[`' is not in
thederivatives table
# This works
Dear list,
I would like to use this C call:
tmp - .Call(R_MPinv, covariance, tol, svdmem)
but it gives me the following error:
Error in getClass(LinStatExpectCovarMPinv) :
LinStatExpectCovarMPinv is not a defined class
I think the reason for this is that the C code starts with this
Dear list members,
sorry for my incompleteness!
My problem is the following:
(R 2.3.1 on Mac OS X 10.4.7 RAM 1GByte using Mac GUI)
I have a function like this:
foo1 - function()
{
out - NULL
for(i in 1:10010) out - paste(out, i, . line\n, sep=)
return(out)
}
a -
G'day all.
I'm a new user of R -- but an arms-length user, as I'm running it from
Octave via the ROctave interface that Duncan Temple Lang wrote some
years ago and Stephan Van Der Walt recently updated for use with Octave
2.1.71. I'm using R version 2.1.1. ROctave uses libR.so to provide the
Dear all,
My question is concerned to the kind how a function is called.
Example A:
foo(1)
Example B:
a - foo(1)
Is there any way for the function foo() to recognise whether the
returned value of foo() is stored in a variable or not, i.e. to
distinguish between Example A and B?
R has a command-line option to set the ppstack size,
--max-ppsize=NSet max size of protect stack to N
Looks like you need to supply this (and it can be done with embedded R)
if the problem persists with current R.
You could also try arima0 or even ar to do the fitting.
On Tue, 15 Aug
Thanks a lot, Ted, for the good extensive explanation.
Let me try summarize, confirm and add a bit w/o repeating too much:
- If you get regular mailing list e-mails, and reply to
postings, any decent mail software will take the
'Message-Id:' header of the message you reply to, and
produce
The pro's and con's of using scale breaks were discussed by
Cleveland (1985) The Elements of Graphing Data (Wadsworth, pp. 85-91,
149). I don't know what Cleveland said about this is the second edition
Spencer Graves:
but I believe there are times when scale breaks are
appropriate,
Dear R Users,
How to make a plot with a bisector is my question. Indeed, I want to plot a
QQplot. But it doesn't have such a bisector.
Thanks a lot.
Amir
-
[[alternative HTML version deleted]]
Hi all.
I have a function that I would like to use either the argument name as
originally defined or another name. Within the function (and other functions)
use the argument name as originally written, so I don't want to simply remove
the old argument name for the new one, but simply allow
foo - function(arg1, this, that) {
if (missing(this) !missing(that)) this - that
if(this 0) stop(this must be positive)
return(arg1/this)
}
foo(arg1=5, this=10)
foo(arg1=5, that=10)
__
R-help@stat.math.ethz.ch mailing list
Is there anyway to change any y[i] value (i=2,...6) to make following NLS
workable?
x - c(0,5,10,15,20,25,30)
y - c(1.0,0.82000,0.68000,0.64000,0.7,0.68667,0.64000)
lm(1/y ~~ x)
nls(1/y ~~ a+b*x^c, start=list(a=1.16122,b=0.01565,c=1), trace=TRUE)
#0.0920573 : 1.16122
You problem is x^c for x = 0. If you intended only c 1, try a starting
value meeting that condition (but it seems that the optimal c is about
0.27 is you increase x slightly).
Why have you used ~~ ? (Maybe because despite being asked not to, you
sent HTML mail?)
On Tue, 15 Aug 2006,
foo - function(arg1, this=that, that)
...
works.
On Tue, 15 Aug 2006, [EMAIL PROTECTED] wrote:
Hi all.
I have a function that I would like to use either the argument name as
originally defined or another name. Within the function (and other functions)
use the argument name as
Hi
I was wondering if you could tell me how would I go abt fitting the following
model to a data series. It is an ARCH jump diffusion model.
X(t)=k+ ó (t)z(t)+J x Y, and
ó(t)= sqrt[a +b [X(t-1)-k]^2]
Here k is mean of the series.
z(t) is iid standard normal error
J is a jump variable
Prof Brian Ripley [EMAIL PROTECTED] writes:
You problem is x^c for x = 0. If you intended only c 1, try a starting
value meeting that condition (but it seems that the optimal c is about
0.27 is you increase x slightly).
Surely you mean c 0.
nls(1/y ~ a+b*x^exp(c),
Hi
Why do you want to change your variable values? It smells a rat to
me.
If you just change your a,b,c values nls arrives to some finite
result (e.g. c=1.5 or c=0.3) . BTW by what magic you obtained such
precise and wrong estimates for a,b,c?
HTH
Petr
On 15 Aug 2006 at 5:54, Xiaodong
On Tue, 15 Aug 2006, Peter Dalgaard wrote:
Prof Brian Ripley [EMAIL PROTECTED] writes:
You problem is x^c for x = 0. If you intended only c 1, try a starting
value meeting that condition (but it seems that the optimal c is about
0.27 is you increase x slightly).
Surely you mean c
Is there a way to get the following code to include
NA values where the coefficients are NA?
((summary(reg))$coefficients)
explanation:
Using a loop, I am running regressions on several
subsets of data1.
reg - ( lm(lm(data1[,1] ~., data1[,2:l])) )
My regression has 10 independent
Consider this situation:
x1 - runif(100); x2 - runif(100); y - 2 + 3*x1 - 4*x2 + rnorm(100)
m1 - summary(lm(y ~ x1))
m2 - summary(lm(y ~ x2))
m3 - summary(lm(y ~ x1 + x2))
Now you have estimated 3 different competing models, and suppose you
want to present the set of models in one table.
Hi folks!
Does anybody know how to use the OLS function in fMultivar to do dynamic
regression? I've tried specifying lags in OLS using a data series
created in fSeries and it doesn't seem to work. I've done dynamic
regression using dyn$lm and I was wondering how to accomplish the same
thing
Hi
On 15 Aug 2006 at 7:01, r user wrote:
Date sent: Tue, 15 Aug 2006 07:01:13 -0700 (PDT)
From: r user [EMAIL PROTECTED]
To: rhelp r-help@stat.math.ethz.ch
Subject:[R] question re: summarry.lm and NA values
Is there a way to
Hi
just as a quick workaround you probably can use aliased value from
summary
fff-rep(summary(reg)$aliased,4)
dim(fff)-c(no.of.your.variables,4)
fff[which(fff)]-NA
fff[which(!fff)]-coef(summary(reg))
to get coef matrix with NA values
HTH
Petr
On 15 Aug 2006 at 17:15, r user [EMAIL
Hello list,
A colleague of mine mentioned a great course on Regression Modeling
Strategies in R. Anyone knows if this course is offered as public course in DC
area?
Thanks a bunch - TM
-
[[alternative HTML version deleted]]
Hi everyone,
I have been using REML to derive intercepts and coeficients for each
individual in a growth study. So the code is
m2 - lmer(change.wt ~ newwt+(newwt|id), data = grow)
Calling coef(model.lmer) gives a matrix with this information which is
what I want. However, as a test I looked at
On Tue, 15 Aug 2006, Petr Pikal wrote:
Hi
On 15 Aug 2006 at 7:01, r user wrote:
Date sent:Tue, 15 Aug 2006 07:01:13 -0700 (PDT)
From: r user [EMAIL PROTECTED]
To: rhelp r-help@stat.math.ethz.ch
Subject: [R] question re:
Is there a way to... always has the answer yes in R (or C or any
language for that matter). The question is: Is there a GOOD way...? where
good depends on the specifics of the situation. So after that polemic,
below is an effort to answer, (adding to what Petr Pikal already said):
-- Bert Gunter
I don't this is because you are using REML. The BLUPs from a mixed model
experience some shrinkage whereas the OLS estimates would not.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Simon Pickett
Sent: Tuesday, August 15, 2006 11:34 AM
To:
I'm having problems getting nls to agree that convergence has occurred in a
toy problem.
nls.out never gets defined when there is an error in nls. Reaching the
maximum number of iterations is alway an error, so nls.out never gets
defined when the maximum number of iterations is reched.
From
Hi all,
Suppose I have a data frame myDF, col A is factor, col B is numeric, col C
is character. I can get their classes by
class(myDF$A)
but is there a quick way to show what classes of all columns are? Thank you.
Tian
[[alternative HTML version deleted]]
On Tue, 2006-08-15 at 13:10 -0400, T Mu wrote:
Hi all,
Suppose I have a data frame myDF, col A is factor, col B is numeric, col C
is character. I can get their classes by
class(myDF$A)
but is there a quick way to show what classes of all columns are? Thank you.
Tian
Depending upon
Hi list,
I'm trying to limit the z axis plotted using persp3d {rgl} to a given
range. I'm trying the following statement
persp3d(g0,pa,D, xlim=range(g0),ylim=range(pa), zlim=range
(zbounds),col=lightblue)
where g0 and pa are both a range of 20 numbers from 0.05 to 1, and D
is a 20 x 20
I am using rexp to generate several exponential distributions. I am passing
rexp a vector of rates , r. I am wanting to simulate a sample of size 200
for each rate so the code looks like: rexp(n=200*length(r),rate=r) this
gives me a vector of the random exponential variables, but they are all
I tried this over a the grasp users yahoo group and got no
responseSo I wonder if anyone here knows about grasper
I keep getting this error when trying to run a model.
Error in smooth.construct.tp.smooth.spec(object, data, knots) :
Too many knots for t.p.r.s term: see `gam.control' to
Try:
rexp(n=200*length(r), rate=rep(r, each=200) )
Hope this helps,
--
Gregory (Greg) L. Snow Ph.D.
Statistical Data Center
Intermountain Healthcare
[EMAIL PROTECTED]
(801) 408-8111
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Spencer Jones
Earl F. Glynn efg at stowers-institute.org writes:
Here's my toy problem:
?nls.control
?nls
# Method 2
X - 0:15
Y - 9.452 * exp(-0.109*X) + 5.111 # Toy problem
nls.out - nls(Y ~ a*exp(b*X)+c,
+start=list(a=6,b=-0.5,c=1),
+
Hi,
I'm new to R; this is my second email to this forum. I have a dataset that
looks like this:
GeneA B
C D
A 511 6116
15151515
B
Hi all,
I got an error message if I set model =F in polr(), like
polr(y ~ x1 + x2, data1, model = F, method = probit)
Error in model.frame(formula, rownames, variables, varnames, extras,
extranames, :
variable lengths differ (found for '(model)')
but
polr(y ~ x1 + x2, data1, method
Try ``model = FALSE'' rather than ``model = F'' and see if it makes a
difference. You make have an unwanted variable named ``F'' lurking
somewhere.
(In general it is a *bad* idea to use ``F'' when ``FALSE'' is
intended.)
cheers,
Rolf
Dieter Menne [EMAIL PROTECTED] wrote in message
news:[EMAIL PROTECTED]
Earl F. Glynn efg at stowers-institute.org writes:
This toy problem is exactly what the warning is for:
Warning
Do not use nls on artificial zero-residual data.
Add some noise and try again.
Thank you!
I had adapted
Hi,
I'm new to R; this is my second email to this forum. I have a dataset that
looks like this:
Gene AB
12536.25 2532.2
22527.35 4583.3
There are about 4 genes with 3-10 patients per dataset. Gene expression
values fill
Or, maybe there's something I don't understand about the
algorithm being used.
Indeed! So before making such comments, why don't you try to learn about it?
Doug Bates is a pretty smart guy, and I think you do him a disservice when
you assume that he somehow overlooked something that he
Hi dear R community,
I have up to 12 measures of a protein for each of 6 patients, taken
every two or three days. The pattern of the protein looks periodic,
but the height of the peaks is highly variable. I'm testing for
periodicity using a Monte Carlo simulation envelope approach applied
to a
Berton Gunter [EMAIL PROTECTED] wrote in message
news:[EMAIL PROTECTED]
Or, maybe there's something I don't understand about the
algorithm being used.
Indeed! So before making such comments, why don't you try to learn about
it?
Doug Bates is a pretty smart guy, and I think you do him a
Hi all,
I am using polr(). The resulting coefficients of first levels are always 0.
What to do if I wnat to get the coefficients of the last level 0.
For example, suppose x has 3 levels, 1, 2, 3
probit - plor(y ~ x, data1, method='probit')
will get coefficients of level 2, 3 of x, but I want
Tim McDonald wrote:
Hello list,
A colleague of mine mentioned a great course on Regression Modeling
Strategies in R. Anyone knows if this course is offered as public course in
DC area?
Thanks a bunch - TM
I'll be teaching this course in DC Sept 28-29 for XL Solutions.
you can use _relevel_ to Reorder Levels of Factor.
use ?relevel to get more information.
2006/8/16, T Mu [EMAIL PROTECTED]:
Hi all,
I am using polr(). The resulting coefficients of first levels are always 0.
What to do if I wnat to get the coefficients of the last level 0.
For example,
I'm using specify.model for the sem package. I can't figure out how to
represent the residual errors for the observed variables for a CFA
model. (Once I get this working I need to add some further constraints.)
Here is what I've tried:
model.sa - specify.model()
F1 - X1,l11, NA
F1 -
49 matches
Mail list logo