gmane.comp.lang.r.general

scale change

Preetam Pal — 2013-05-21T06:00:47

Hi all,

I am sure this is simple, but being a beginner in R,I am finding it
difficult to manage myself.

I am running the following code to get a multiple line plot for two time
series variables, x and y:
plot(as.ts(cbind(x,y)), plot.type = "single", col = 1:2)

I want to change the scales of both axes:
I need to 'stretch' the x-axis and 'shrink' the y-axis (1/2cm on the
previous x-axis would now need to be stretched to  1 cm on the new one and
exactly the opposite for the y-axis).

I am not being able to do this.
Any help is welcome.

ordered and unordered variables

meng — 2013-05-21T05:35:59

Hi all:
If the explainary variables are ordinal,the result of regression is different from
"unordered variables".But I can't understand the result of regression from "ordered
variable".
 
The data is warpbreaks,which belongs to R.
 
If I use the "unordered variable"(tension):Levels: L M H
The result is easy to understand:
    Estimate Std. Error t value Pr(>|t|)   
(Intercept)    36.39       2.80  12.995  < 2e-16 ***
tensionM      -10.00       3.96  -2.525 0.014717 * 
tensionH      -14.72       3.96  -3.718 0.000501 ***

If I use the "ordered variable"(tension):Levels: L < M < H
I don't know how to explain the result:
           Estimate Std. Error t value Pr(>|t|)   
(Intercept)   28.148      1.617  17.410  < 2e-16 ***
tension.L    -10.410      2.800  -3.718 0.000501 ***
tension.Q      2.155      2.800   0.769 0.445182   
 
What's "tension.L" and "tension.Q" stands for?And how to explain the result then?
 
Many thanks.
 

 
[[alternative HTML version deleted]]

Re: How many decimal places of information does R actually usein computation?

Bert Gunter — 2013-05-21T03:39:34

See, e.g.

http://en.wikipedia.org/wiki/Double-precision_floating-point_format

Also in ?print.default  please note:

Large number of digits

Note that for large values of digits, currently for digits >= 16, the
calculation of the number of significant digits will depend on the
platform's internal (C library) implementation of sprintf()
functionality.

How many decimal places of information does R actually use incomputation?

Xiao He — 2013-05-21T01:06:16

Dear R-users:

Hi, I read here (
http://stackoverflow.com/questions/2287616/controlling-digits-in-r) that R
is only accurate up to the 15th decimal place, despite the fact that if you
choose to display more decimal places, it will. I wonder if R uses the
information beyond the 15th decimal place in actual computation. Thanks!


Best,
Xiao

[[alternative HTML version deleted]]

Re: how to GREP out a string like this......THANKS.

arun — 2013-05-21T01:48:16

Hi,
May be this helps.

lines<- readLines(textConnection("NM_019397 // Egfl6 // EGF-like-domain, multiple 6 // X F5|X 71.5 cM // 54156
NM_019397 // Egfl7 // EGF-like-domain, multiple 6 // X F5|X 71.5 cM // 54158"))
library(stringr)
word(lines,2,sep=" // ")
#[1] "Egfl6" "Egfl7"

lines1<- readLines(textConnection("NM_019397 // Egfl6 // EGF-like-domain, multiple 6 // X F5|X 71.5 cM // 54156
NM_019397 // Egfl7 domain // EGF-like-domain, multiple 6 // X F5|X 71.5 cM // 54158"))
 word(lines1,2,sep=" // ")
#[1] "Egfl6"        "Egfl7 domain"
A.K.



----- Original Message -----
From: Hon Kit (Stephen) Wong <honkit< at >stanford.edu>
To: r-help< at >r-project.org
Cc: 
Sent: Monday, May 20, 2013 7:45 PM
Subject: [R] how to GREP out a string like this......THANKS.

Dear ALl,

I hope you could help me out on this simple problem. I have many thousand lines like this:
NM_019397 // Egfl6 // EGF-like-domain, multiple 6 // X F5|X 71.5 cM // 54156

I want to extract the string inside the first // //, in this case is Egf16. 


How do I apply grep function?

Thanks.

Stephen HK Wong

Stephen HK Wong
Research Associate,Cleary Lab
Lab Phone: 650-723-5340
MC 5457 
Lokey Stem Cell Research Building 
265 Campus Drive, Rm. G2035 
Stanford, California 94305-5324

______________________________________________
R-help< at >r-project.org mailing list
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-contained, reproducible code.

Re: how to GREP out a string like this......THANKS.

William Dunlap — 2013-05-21T00:54:59

You suggested
  > lapply( lines, function(l) strsplit(l, " // ")[[1]][2] )

strsplit is vectorized so the following is equivalent but simpler and quicker:
   lapply(strsplit(lines, " // "), function(x)x[2])

The OP probably wants a character vector, not a list so use sapply or vapply (safer
than sapply and a bit quicker).  Any of the following would do:
  vapply(strsplit(lines, " // "), `[`, 2, FUN.VALUE="")
  vapply(strsplit(lines, " // "), function(x)x[2], FUN.VALUE="")
  sapply(strsplit(lines, " // "), `[`, 2) # wrong answer if length(lines)==0


Bill Dunlap
Spotfire, TIBCO Software
wdunlap tibco.com

Re: how to GREP out a string like this......THANKS.

David Winsemius — 2013-05-21T00:17:02

On May 20, 2013, at 4:45 PM, Hon Kit (Stephen) Wong wrote:


[1] "Egfl6"

You can use;

lapply( lines, function(l) strsplit(l, " // ")[[1]][2] )



Well, grep is only going to give you a test and you want a replacement or extraction function. sub or gsub would be possibilities but they are greedy so its a bit more difficult to constrain their targeting to only the first  and second "//".

how to GREP out a string like this......THANKS.

Hon Kit (Stephen) Wong — 2013-05-20T23:45:31

Dear ALl,

I hope you could help me out on this simple problem. I have many thousand lines like this:
NM_019397 // Egfl6 // EGF-like-domain, multiple 6 // X F5|X 71.5 cM // 54156

I want to extract the string inside the first // //, in this case is Egf16. 


How do I apply grep function?

Thanks.

Stephen HK Wong

Stephen HK Wong
Research Associate,Cleary Lab
Lab Phone: 650-723-5340
MC 5457 
Lokey Stem Cell Research Building 
265 Campus Drive, Rm. G2035 
Stanford, California 94305-5324

Re: Gamma curve fit to data with specific bins

Rui Barradas — 2013-05-20T22:49:13

Hello,

You are fitting a vector other than the vector 'x'.
And you are mistaking the parameter scale for rate.

est <- fitdistr(x,"gamma")$estimate


#plot the gamma curve with the found parameters
hist(x, breaks=Size, freq=FALSE, xlab="Drop Size", ylab="No. of Drops")
curve(dgamma(x, rate=est["rate"], shape=est["shape"]),from=0, to=16, 
main="Gamma
distribution", ylab="Probability", add = TRUE, col = "red")


Hope this helps,

Rui Barradas

Em 20-05-2013 16:44, Lorentz, Andrew escreveu:

Re: bar plot with non-zero starting level

Jim Lemon — 2013-05-20T22:40:34

Hi Xianwen,
The problem with the overlapping is in the last two lines. These should 
read:

         geom_rect(data = rect_MNL_Delta, aes(xmin = 
rect_MNL_Delta$xmin, xmax = rect_MNL_Delta$xmax, ymin = 
rect_MNL_Delta$ymin, ymax = rect_MNL_Delta$ymax), fill = "blue", alpha = 
0.1) +
         geom_rect(data = rect_GMNL, aes(xmin = rect_GMNL$xmin, xmax = 
rect_GMNL$xmax, ymin = rect_GMNL$ymin, ymax = rect_GMNL$ymax), fill = 
"green", alpha = 0.1)

I am not terribly familiar with ggplot, but you should be able to get 
custom tick labels on your horizontal axis. I have to go out now, but 
when I return I'll have a look at the "scale" functions, as I think this 
is hiding in there.

Jim

Re: as.vector with mode="list" and POSIXct

Jeff Newmiller — 2013-05-20T22:39:53

I don't know what you plan to do with this list, but lists are quite a bit less efficient than fixed-mode vectors, so you are likely losing a lot of computational speed by using this list. I don't hesitate to use simple data frames (lists of vectors), but processing lists is on par with for loops, not vectorized computation. It may still support a simpler model of computation, but that is an analyst comprehension benefit rather than a computational efficiency benefit.
---------------------------------------------------------------------------
Jeff Newmiller                        The     .....       .....  Go Live...
DCN:<jdnewmil< at >dcn.davis.ca.us>        Basics: ##.#.       ##.#.  Live Go...
                                      Live:   OO#.. Dead: OO#..  Playing
Research Engineer (Solar/Batteries            O.O#.       #.O#.  with
/Software/Embedded Controllers)               .OO#.       .OO#.  rocks...1k
--------------------------------------------------------------------------- 
Sent from my phone. Please excuse my brevity.

Alexandre Sieira <alexandre.sieira< at >gmail.com> wrote:


______________________________________________
R-help< at >r-project.org mailing list
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-contained, reproducible code.

Re: stack object layer names not visible

Uwe Ligges — 2013-05-20T22:26:37


On 15.05.2013 17:42, Fabio Berzaghi wrote:


Perhaps the package got updated?

Best,
Uwe Ligges

Re: old Windows binary for GBM package

Uwe Ligges — 2013-05-20T22:20:46


On 20.05.2013 23:24, Andrew Z wrote:

At least CRAN does not keep copies of old binaries, there is a version 
2.0-8 for R-2.14.x, and version 1.6-3.1 for R-2.13.x, but hese R 
versions are rather ancient, hence I'd suggest to try to learn how to 
install from sources, in most cases it is not too hard.

Or ask the maintainer for a fix again so that we see a proper update on 
CRAN soon.

Best,
Uwe Ligges

Re: old Windows binary for GBM package

David Winsemius — 2013-05-20T22:17:26

On May 20, 2013, at 2:24 PM, Andrew Z wrote:


Several mirrors have old binaries. This happens to be the one closest to me:

http://cran.cnr.berkeley.edu/web/packages/

http://cran.cnr.berkeley.edu/bin/windows/contrib

http://cran.cnr.berkeley.edu/bin/windows/contrib/2.14/gbm_2.0-8.zip

Go Bears!

Re: R for windows GUI front-end has stopped working

Uwe Ligges — 2013-05-20T22:15:50


On 20.05.2013 18:10, Naser Jamil wrote:

A crash is always a bug, but now the question is where:

What is Windows 2007?

Which R version are we talking about? R-3.0.1 or R-devel?

Where is reproducible code?

Is there any package or self written external code involved in which 
case it may be not base R that causes the error.

Please re-read the posting guide that suggests how to ask questions so 
that we are able to help.

Best,
Uwe Ligges

old Windows binary for GBM package

Andrew Z — 2013-05-20T21:24:50

Is there a place to an old version (GBM 1.6 or 2.0) of the Windows
64-bit binary for the GBM package?

In GBM 2.1, CV does not work on any of my data sets, so I reported it
to https://code.google.com/p/gradientboostedmodels/ . However, I would
like to soon continue to use GBM for a project, and I prefer not to
set up a build environment on Windows to build from source.


Andrew

Re: as.vector with mode="list" and POSIXct

Alexandre Sieira — 2013-05-20T21:39:30

So it does. I had tried list() but not as.list(). Silly me. :)

Thank you very much, William.

-- 
Alexandre Sieira
CISA, CISSP, ISO 27001 Lead Auditor

"The truth is rarely pure and never simple."
Oscar Wilde, The Importance of Being Earnest, 1895, Act I
On 20 de maio de 2013 at 18:32:35, William Dunlap (wdunlap< at >tibco.com) wrote:
Try using as.list(x) instead of as.vector(x, mode="list").  
The former has a method for POSIXct; the latter does not.  

[1] "2013-05-20 14:28:00 PDT" "2013-11-30 22:10:00 PST"  
[[1]]  
[1] "2013-05-20 14:28:00 PDT"  

[[2]]  
[1] "2013-11-30 22:10:00 PST"  

Bill Dunlap  
Spotfire, TIBCO Software  
wdunlap tibco.com  


> Oscar Wilde, The Importance of Being Earnest, 1895, Act I

Re: as.vector with mode="list" and POSIXct

William Dunlap — 2013-05-20T21:32:32

Try using as.list(x) instead of as.vector(x, mode="list").
The former has a method for POSIXct; the latter does not.

  > x <- as.POSIXct(c("2013-05-20 14:28", "2013-11-30 22:10"), tz="US/Pacific") 
  > x
  [1] "2013-05-20 14:28:00 PDT" "2013-11-30 22:10:00 PST"
  > as.list(x)
  [[1]]
  [1] "2013-05-20 14:28:00 PDT"
  
  [[2]]
  [1] "2013-11-30 22:10:00 PST"

Bill Dunlap
Spotfire, TIBCO Software
wdunlap tibco.com


______________________________________________
R-help< at >r-project.org mailing list
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-contained, reproducible code.

as.vector with mode="list" and POSIXct

Alexandre Sieira — 2013-05-20T21:09:56

I was trying to convert a vector of POSIXct into a list of POSIXct, However, I had a problem that I wanted to share with you.

Works fine with, say, numeric:


[1] 1 2 3
 num [1:3] 1 2 3
[[1]]
[1] 1

[[2]]
[1] 2

[[3]]
[1] 3

List of 3
 $ : num 1
 $ : num 2
 $ : num 3

If you try it with POSIXct, on the other hand…


[1] "2013-05-20 18:02:07 BRT" "2013-05-20 18:02:07 BRT"
 POSIXct[1:2], format: "2013-05-20 18:02:07" "2013-05-20 18:02:07"
[[1]]
[1] 1369083728

[[2]]
[1] 1369083728

List of 2
 $ : num 1.37e+09
 $ : num 1.37e+09

The POSIXct values are coerced to numeric, which is unexpected.

The documentation for as.vector says: "The default method handles 24 input types and 12 values of type: the details of most coercions are undocumented and subject to change." It would appear that treatment for POSIXct is either missing or needs adjustment.

Unlist (for the reverse) is documented to converting to base types, so I can't complain. Just wanted to share that I ended up giving up on vectorization and writing the two following functions:


unlistPOSIXct <- function(x) {
  retval = rep(Sys.time(), length(x))
  for (i in 1:length(x)) retval[i] = x[[i]]
  return(retval)
}

listPOSIXct <- function(x) {
  retval = list()
  for (i in 1:length(x)) retval[[i]] = x[i]
  return(retval)
}

Is there a better way to do this (other than using *apply instead of for above) that better leverages vectorization? Am I missing something here?

Thanks!




-- 
Alexandre Sieira
CISA, CISSP, ISO 27001 Lead Auditor

"The truth is rarely pure and never simple."
Oscar Wilde, The Importance of Being Earnest, 1895, Act I

Re: help with 'cem' for r 2.14.2

Duncan Murdoch — 2013-05-20T20:51:47

I think you need to talk to SPSS support for this.  We have no 
connection with SPSS here.

As well, the current release of R is 3.0.1.  R 2.14.2 is about 15 months 
old; that's ancient history as far as we're concerned (though it may be 
cutting edge for SPSS).

Duncan Murdoch


I have run a test for R and it appears to be running correctly.  I then 
downloaded psmatching3 and have tried to use the PS matching dialog in 
SPSS.  However, I continue to run into problems as SPSS reports that 
there is no 'cem' file.  I have tried to download cem separately from 
your website (both the new and the older version) and to install it 
either in R or as an extension in SPSS but I receive the following errors:

Re: Generate positive definite matrix with constraints

Gabor Grothendieck — 2013-05-20T20:04:46

provide an m0 matrix of the appropriate dimensions and change 2 to the
new dimension in the relevant lines.


--
Statistics & Software Consulting
GKX Group, GKX Associates Inc.
tel: 1-877-GKX-GROUP
email: ggrothendieck at gmail.com