gmane.comp.lang.r.general

Re: R Error: System is computationally singular

Nathan Svoboda — 2012-05-24T19:17:35

Thank you for your quick reply, 

When I run the code you provide I get this output:

LCOVER
LOCS      1      2      3      4      5      6      7      9
   0 214507  79939  69803 778359  22932  32391  99630   8082
   1     15      7      1     32      0      0      0      0
   2      2      1      0      0      0      0      0      0
   3      0      0      0      1      0      0      0      0

Nate



--
View this message in context: http://r.789695.n4.nabble.com/R-Error-System-is-computationally-singular-tp4631242p4631251.html
Sent from the R help mailing list archive at Nabble.com.

Re: R Error: System is computationally singular

Nathan Svoboda — 2012-05-24T19:16:23

Thank you for your quick reply! 
 
When I run the code you provide I get this output:
 
LCOVER
LOCS      1      2      3      4      5      6      7      9
   0 214507  79939  69803 778359  22932  32391  99630   8082
   1     15      7      1     32      0      0      0      0
   2      2      1      0      0      0      0      0      0
   3      0      0      0      1      0      0      0      0
 
Nate
 
 
 

________________________________

From: David Winsemius [mailto:dwinsemius< at >comcast.net]
Sent: Thu 5/24/2012 1:54 PM
To: Nathan Svoboda
Cc: r-help< at >r-project.org
Subject: Re: [R] R Error: System is computationally singular




On May 24, 2012, at 1:57 PM, Nathan Svoboda wrote:


What does this show:

with( FAWNS, table(LOCS, LCOVER) )

--
David Winsemius, MD
West Hartford, CT




[[alternative HTML version deleted]]

Re: R Error: System is computationally singular

David Winsemius — 2012-05-24T18:54:55

On May 24, 2012, at 1:57 PM, Nathan Svoboda wrote:


What does this show:

with( FAWNS, table(LOCS, LCOVER) )

R Error: System is computationally singular

Nathan Svoboda — 2012-05-24T17:57:02

Greetings,

I am trying to fit a zero-inflated Poisson model using zeroinfl() from the
pscl library. I have 5 covariates (4 continuous, 1 categorical); the
categorical variable has 7 levels.  I have had success fitting models that
contain only the continuous covariates; however, when I add the categorical
variable to any of the models (or if I run it by itself) I get the following
error: 

Error in solve.default(as.matrix(fit$hessian)) : 

  system is computationally singular: reciprocal condition number =
3.46934e-20

 The code I am using is:

library(pscl)
f1 <- formula(LOCS ~ as.factor(LCOVER) + D_ROADS + D_WATER + D_EDGE +
D_GRASS)
ZIP1 <- zeroinfl(f1, dist="poisson", link = "logit", data = FAWNS)
 
There is no correlation between my covariates. Also, I tried reducing my
categorical covariate to 3 levels and still receive the same error. Can
anyone suggest why I may be getting this error when I add the categorical
covariate?
 
I appreciate your time and input. Thank you,
 
Nate
 
Nathan Svoboda
Graduate Research Assistant
Carnivore Ecology Lab
Mississippi State University

--
View this message in context: http://r.789695.n4.nabble.com/R-Error-System-is-computationally-singular-tp4631242.html
Sent from the R help mailing list archive at Nabble.com.

Re: R does not recognise columns and rows as they are supposed to be

William Dunlap — 2012-05-24T18:26:01

Note that 2-dimensional subscripts are given as
    mat[ROWS, COLUMNS]
and you are using the reverse order, hence you get the
error message that you are asking for rows 364:369 of
a 360-row matrix.


The file may well contain single precision numbers and using
  what="double", size=4
will read them and convert them to doubles so other functions
in R can use them.  (Likewise, readBin can read files containing
1- or 2-byte signed or unsigned integers and store them as the
4-byte signed integers that the rest of R can deal with.)

If the file was created on another sort of machine, then you may
have worry about the byte order - try adding endian="big" or
endian="little".

Re: applying cbind (or any function) across all components in alist

David Winsemius — 2012-05-24T18:25:59

On May 24, 2012, at 12:49 PM, Hans Thompson wrote:


Well, it's not clear what "function" you want applied to "what". In  
fact, it's even less clear in this posting than it was in your first  
one (since you assume, incorrectly, that we are looking at this on  
Nabble. Readers of this list expect you to include context.)  I could  
not figure out in your first posting how we should be considering the  
fourth expression

   {(l1[[1]][,3]+l2[[1]][,2])/2 }

... to be an "average" of anything since its indices do not match?  
Perhaps equivalently, why should a 2 x 3 structure merged <somehow> to  
a 2 x 2 structure yield a 2 x 4 structure?

Re: plot(summary) quantreg - Not all outputs needed

Peter Ehlers — 2012-05-24T18:10:57

Isn't this accomplished by specifying the argument 'parm'?

   plot(fit, parm = 2)

etc.

Peter Ehlers

Re: applying cbind (or any function) across all components inalist

David L Carlson — 2012-05-24T18:10:14

You don't make it clear if you want the two lists processed in parallel, but
if so, this may be a step toward the function you want:

Combine <- function(l1, l2) {
  pattern <- cbind(c(1, 2, 2, 3), c(1, 1, 2, 2))
  for (i in 1:length(l1)) {
    print((l1[[i]][,pattern[,1]]+l2[[i]][,pattern[,2]])/2)
  }
}
Combine(l1, l2)

This simply prints the results, but you didn't indicate how you wanted the
output organized.

----------------------------------------------
David L Carlson
Associate Professor of Anthropology
Texas A&M University
College Station, TX 77843-4352

Re: R does not recognise columns and rows as they are supposed to be

Jonsson — 2012-05-24T17:56:28

Yes this helped a lot .
I exactly followed what you suggested:
X<-(82:92) ; Y<-(364:369) #####   for sellected region
+       conne <- file(listfile[i], "rb")
+       file1<- readBin(conne, double(),  n=360*720)
+      file2<-matrix(data=file1,ncol=720,nrow=360)
+      extract[i]<-mean(file2[X,Y],na.rm=TRUE)
+       close(conne)
+ write.table(extract,"C:\\Users\\aalyaari\\Desktop\\New folder
(10)\\samregion1.txt")}

But I wonder why  I got all values like

-3.75E+306
-1.30E+54
-1.22E+58
and the right ones should be like:
22.25
22.76
33.25



--
View this message in context: http://r.789695.n4.nabble.com/R-does-not-recognise-columns-and-rows-as-they-are-supposed-to-be-tp4631217p4631241.html
Sent from the R help mailing list archive at Nabble.com.

Re: inner_perc_table?

Charles Determan Jr — 2012-05-24T17:39:00

Thank you William,

Does this mean I should use "" when inner_perc_table is called in the code
then or am I supposed to define a full path?

Regards,
Charles

On Thu, May 24, 2012 at 12:31 PM, William Dunlap <wdunlap< at >tibco.com> wrote:


[[alternative HTML version deleted]]

Re: inner_perc_table?

William Dunlap — 2012-05-24T17:31:31

You need to supply a character string, e.g., "inner_perc_table", that names
the C entry point.  If you supply a name, e.g., entry_point,  or a more general expression,
e.g, entry_points[j], then it will be evaluated with the usual R evaluation
rules and the result must be a character string.

  > getNativeSymbolInfo("inner_perc_table")
  $name
  [1] "inner_perc_table" 

  $address
  <pointer: 0x000000006b1cb650>
  attr(,"class")
  [1] "NativeSymbol"

  $package
  DLL name: nlme
  Filename: c:/Program Files/R/R-2.15.0/library/nlme/libs/x64/nlme.dll
  Dynamic lookup: FALSE

  $numParameters
  [1] 6

  attr(,"class")
  [1] "CRoutine"         "NativeSymbolInfo"

or
  > entry_point <- "inner_perc_table"
  > entry_points <- c("outer_perc_table", "inner_perc_table")
  > identical(getNativeSymbolInfo(entry_point), getNativeSymbolInfo("inner_perc_table"))
  [1] TRUE
  > identical(getNativeSymbolInfo(entry_points[2]), getNativeSymbolInfo("inner_perc_table"))
  [1] TRUE

Bill Dunlap
Spotfire, TIBCO Software
wdunlap tibco.com

Re: R does not recognise columns and rows as they are supposed to be

Rui Barradas — 2012-05-24T16:53:18

Sorry, forgot the last line in the op code.

write.table(extract, ...) is called every time through the loop, not just
one time after it. Put it after closing '}'.

Rui Barradas



Rui Barradas wrote


--
View this message in context: http://r.789695.n4.nabble.com/R-does-not-recognise-columns-and-rows-as-they-are-supposed-to-be-tp4631217p4631229.html
Sent from the R help mailing list archive at Nabble.com.

Re: R does not recognise columns and rows as they are supposed to be

Rui Barradas — 2012-05-24T16:47:30

Hello,

There are several things with your code, most of which are not errors.

1. X<-c(364:369) ; Y<-c(82:92 and later c(1:365)
Expressions of the form m:n are integer sequences, the c() is not needed.

2. extract<-vector()
First you create the vector, then keep extending it throughout the loop.
This is much slower.
If you know the final length (and type), create it like this:
extract <- numeric(365)   # or double(365)

3. 'dir1' is not used.

4. Now the readBin instruction:
4.1. 'file' is the name of a function, use, say, 'file1' instead.
4.2. Option 'signed' defaults to TRUE but it's only valid for integer types.

5. And, maybe this is the problem, in R, doubles are 8 bytes, not 4. Get rid
of option 'size', like the help page says,

"If size is specified and not the natural size of the object, each element
of the vector is coerced to an appropriate type before being written or as
it is read."

You are NOT reading doubles.

As for the rest, it seems ok to me. (Keep option 'n' in readBin.)

Hope this helps,

Rui Barradas

Jonsson wrote


--
View this message in context: http://r.789695.n4.nabble.com/R-does-not-recognise-columns-and-rows-as-they-are-supposed-to-be-tp4631217p4631227.html
Sent from the R help mailing list archive at Nabble.com.

Re: applying cbind (or any function) across all components in a list

Hans Thompson — 2012-05-24T16:49:13

I'm confused why I haven't made clear what I am asking for help with.  

I have two different lists with two (or many) components, [[1]] and [[2]]. 
One of the list has components with dim=c(2,3) and the other has dim=c(2,2). 
I want to create a new list with components dim=c(2,4) by binding together
the averages of the columns using 


which should put out:

     [,1] [,2] [,3] [,4]
[1,]    1    2    3    4
[2,]    2    3    4    5

my problem is finding out how I can apply this function to all the
components within the list in the same function.

--
View this message in context: http://r.789695.n4.nabble.com/applying-cbind-or-any-function-across-all-components-in-a-list-tp4631128p4631228.html
Sent from the R help mailing list archive at Nabble.com.

Fwd: help needed

QAMAR MUHAMMAD UZAIR — 2012-05-24T16:45:48



      --- segue il messaggio inoltrato (the forwarded 
message follows) ---

Re: inner_perc_table?

Charles Determan Jr — 2012-05-24T17:30:35

My apologies, feel a little foolish.  I forgot the "" for the object.  This
worked, however, once I have this information, I am not sure what I should
be using to fix the issue.  I can determine the location in  the nlme.dll
file but how can I use this information to have the other function work?

Regards,
Charles

On Thu, May 24, 2012 at 11:51 AM, Duncan Murdoch
<murdoch.duncan< at >gmail.com>wrote:


[[alternative HTML version deleted]]

Re: set tkscale by tkentry

j verzani — 2012-05-24T17:27:22


Greg is right. You might try validating on focusout, rather than the key, 
but this is easy enough to do in R code, rather than let tcl do that work:

a <- 306870; b <- 3026741

tt<-tktoplevel()
varalpha <- tclVar(a)
charalpha <- tclVar(as.character(a))


scale <- tkscale(tt, from=a, to=b, resolution=1, label="alpha",
              variable=varalpha,
              showvalue=TRUE, orient="horiz")
ed <- tkentry(tt, textvariable=charalpha)

tkpack(ed)
tkpack(scale)

## connect
tkconfigure(scale, command=function(...) {
  tclvalue(charalpha) <- as.character(tclvalue(varalpha))
})

valid_input <- function(...) {
  val <- as.numeric(tclvalue(charalpha))
  if(a <= val & val <= b) {
    message("set to ", val)
    tclvalue(varalpha) <- val
  } else {
    message("not valid...")
    tkfocus(ed)
  }
}

tkbind(ed, "<Return>", valid_input)
tkbind(ed, "<FocusOut>", valid_input)

Re: Exclude when sd=0

David L Carlson — 2012-05-24T17:23:05

Assuming you mean that the column is constant so that the sd is 0

test <- matrix(rnorm(50), nrow=10, ncol=5)
test[,2] <- 1
test[,4] <- 2
# test[,which(apply(test, 2, sd) > 0)] could fail on rounding errors
Test2 <- test[, which(apply(test, 2, sd) > 1e-10)] 


----------------------------------------------
David L Carlson
Associate Professor of Anthropology
Texas A&M University
College Station, TX 77843-4352

Re: inner_perc_table?

Charles Determan Jr — 2012-05-24T17:02:01

Thank you Duncan,

I used the R function getNativeSymbolInfo(inner_perc_table) and the object
is not found:

Error in is.vector(X) : object 'inner_perc_table' not found
Done.

I also tried the defining the directory to the nlme.dll file and it had the
same error.
Is there something I missed while doing this?

Thanks,
Charles


On Thu, May 24, 2012 at 11:51 AM, Duncan Murdoch
<murdoch.duncan< at >gmail.com>wrote:


[[alternative HTML version deleted]]

Re: inner_perc_table?

Duncan Murdoch — 2012-05-24T16:51:41

Look for the C function R_registerRoutines and/or the R function 
getNativeSymbolInfo.

Duncan Murdoch

transform 1 col to 2 col

arun — 2012-05-24T16:23:20

Hi,

Just an addendum.

Suppose if you want to add more columns and do the split, you can try,

inscrutable.df2<-data.frame(inscrutable.df1,V2=(c(rnorm(1),"GTDF","SQ:1234","DFFD","DFDSE")[rep(c(1,2,3,4,5),times=4)]))
inscrutable3.df3<-data.frame(matrix(inscrutable.df1$V1,ncol=2,byrow=TRUE),matrix(inscrutable.df2$V2,ncol=2,byrow=TRUE))  

A.K.



----- Original Message -----
From: "MacQueen, Don" <macqueen1< at >llnl.gov>
To: Soheila Khodakarim <lkhodakarim< at >gmail.com>
Cc: "r-help< at >r-project.org" <r-help< at >r-project.org>
Sent: Thursday, May 24, 2012 10:19 AM
Subject: Re: [R] transform 1 col to 2 col

See insertion below