[Rd] R 2.1.1: read.table processes C-style escapes (PR#8037)
In R 2.1.1, the default behaviour of scan() was changed to process all
C-style escapes, even when a delimiter was specified using the 'sep'
argument. A new argument 'allowEscapes' was introduced to turn this
processing off.
Because read.table() calls scan(), read.table() inherits the new default
behaviour of scan() but without a way to turn it off. For example, reading
a file testdata.txt' containing
X
A
\0
C
produces
> read.delim("testdata.txt")
[1] X
<0 rows> (or 0-length row.names)
It seems that all the occurances of scan() within read.table() need to have
'allowEscapes=FALSE' added to the argument string.
Gordon
> version
_
platform i386-pc-mingw32
arch i386
os mingw32
system i386, mingw32
status Patched
major2
minor1.1
year 2005
month07
day 22
language R
-------
Dr Gordon K Smyth, Senior Research Scientist, Bioinformatics,
Walter and Eliza Hall Institute of Medical Research,
1G Royal Parade, Parkville, Vic 3050, Australia
Tel: (03) 9345 2326, Fax (03) 9347 0852,
Email: [EMAIL PROTECTED], www: http://www.statsci.org
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] read.table produces extra rows when file contains extra columns on (PR#9128)
Reading the following delimited file with read.csv() or read.table()
file1:
X,Y
1,2
2,4
3,6
4,8
5,10,,
6,12
produces a data.frame with 7 rows instead of 6 because the two extra values on
line 6 of the file
are pushed into a new row of the data.frame. In other words, the extra columns
on line 6 are
interpreted as a second case on the same line. This contradicts the help
?read.table which states
that cases correspond to lines.
A desirable behaviour might be to ignore the extra columns with a warning. It
would be nice
though to be consistent with the behaviour reading the shorter file
file2:
X,Y
1,2
2,4,,
3,6
which currently produces an error.
Gordon
> read.csv("file1.csv")
X Y
1 1 2
2 2 4
3 3 6
4 4 8
5 5 10
6 NA NA
7 6 12
> read.csv("file2.csv")
Error in read.table(file = file, header = header, sep = sep, quote = quote, :
more columns than column names
> sessionInfo()
Version 2.3.1 (2006-06-01)
i386-pc-mingw32
attached base packages:
[1] "methods" "stats" "graphics" "grDevices" "utils" "datasets"
"base"
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Citation for R
>Note also that R does have a User Guide, i.e., while there is plenty of >excellent documentation, >there is no single document which is a guide to the whole project. Oops, I meant to write "R does not have a User Guide". Just to explain this further, the citation() function asks me to cite a "Manual" with the title "R: A language and environment for statistical computing". Although R comes with excellent documentation, including at least 6 manuals on different aspects of the software, no manual or document with that title actually exists, as far as I know. Gordon __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Citation for R
At 05:22 PM 14/06/2005, [EMAIL PROTECTED] wrote: > >>>>> On Tue, 14 Jun 2005 08:42:59 +1000 (EST), > >>>>> Gordon K Smyth (GKS) wrote: > > > On Tue, June 14, 2005 12:49 am, Thomas Lumley said: > >> On Mon, 13 Jun 2005, Gordon K Smyth wrote: > >> > >>> This is just a note that R would get a lot more citations if the > >>> recommended citation was an article in a recognised journal or from a > >>> recognised publisher. > >>> > >> > >> This is unfortunately true, but R is *not* an article or a book, it is a > >> piece of software. I don't think I'm the only person who thinks it is > >> counterproductive in the long run to encourage users to cite an article > >> that they probably haven't read instead of citing the software they > >> actually used. > >> > >> Jan's suggestion of the Journal of Statistical Software might provide a > >> solution, since JSS *does* publish software. > >> > >> -thomas > > > In the biology world, it is common to publish an article > > announcing a software project, and to cite that. The referees of > > the article are expected to try out and comment on the software. > > This gives the authors credit, and ensures that both the article > > and the software have been peer refereed, at least to a limited > > extent. > >How do you cite books in this world, or to but the question in another >way: How do you make sure a book is peer-reviewd? After all it is >quite easy to become a "publisher" and publish ones own books. Many >university departments I know are registered ISBN publishers >(including our department). Must be hard to distinguish "real" books >from others, I guess. > >Fritz Books are cited as in the statistics literature but, naturally, there is a tendency to prefer references from more reputable sources. Hence a Wiley & Son book would be prefered, other things being equal, to a book from a minor university press, which is turn would be prefered to a self-published book. Self-published electronic books are pretty much at the bottom of the pile. This doesn't mean that important references of this form can't be cited, but it doesn mean that one is pushing uphill. Self-published software manuals are usually cited in-text, as are many other tools and technologies mentioned in the methods section, rather than in the references. Thanks for pointing out, in a separate posting, that citation() refers to the "R Reference Manual". Gordon __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] boxplot() defaults {was "boxplot in extreme cases"}
>[Rd] boxplot() defaults {was "boxplot in extreme cases"}
>Martin Maechler maechler at stat.math.ethz.ch
>Mon Nov 8 10:36:42 CET 2004
>
> AndyL> Try:
>
> AndyL> x <- list(x1=rep(c(0,1,2),c(10,20,40)),
> x2=rep(c(0,1,2),c(10,40,20)))
> AndyL> boxplot(x, pars=list(medpch=20, medcex=3))
>
> AndyL> (Cf ?bxp, pointed to from ?boxplot.)
>
>Good! Thank you, Andy.
>
>However,
>this is not the first time it had crossed my mind that R's
>default settings of drawing boxplot()s are not quite ok -- and
>that's why I've diverted to R-devel.
>
>Keeping Tufte's considerations in mind, (and me not really wanting
>to follow S-plus), shouldn't we consider to slightly change R's
>boxplot()ing such that
>
>boxplot(list(x1=rep(c(0,1,2),c(10,20,40)), x2=rep(c(0,1,2),c(10,40,20
>
>will *not* give too identically looking boxplots?
>Also, the median should be emphasized more by default anyway.
>{The lattice function bwplot() does it by only drawing a large
> black ball as in Andy's example (and not drawing a line at all)}
>
>One possibility I'd see is to use a default 'medlwd = 3'
>either in boxplot() or in bxp(.) and hence, what you currently get by
>
>boxplot(list(x1=rep(c(0,1,2),c(10,20,40)), x2=rep(c(0,1,2),c(10,40,20))),
>medlwd=3)
>
>would become the default plotting in boxplot().
>Of course a smaller value "medlwd=2" would work too, but I'd
>prefer a bit more (3).
>
>Martin
Hi Martin,
I'm not sure this innovation (medlwd=3 default) is a good idea. Boxplots
are designed to display many samples simultaneously on a graph, and it is
important they be as clean and as simple as possible. To my eye, and to
everyone in my lab, the thickened median line is rather distracting and
makes the boxplots look more cluttered ("ugly" one of my postdocs said).
The thickened line also goes against Tufte's principle of using minimum ink
to represent the message.
Yours and Erich's point about distinguishing the median==1st quartile case
from the median==3rd quartile case is well taken. How about making medlwd=3
(or medlwd=2) the default behaviour only when the median coincides with one
of the quartiles? That might satisfy everyone?
I notice that there wasn't any follow up discusssion of this post of the
r-devel list. Did this suggestion get any support? The boxplots have been
so well accepted in their current form for many, many years, decades even,
so one should be especially cautious of making changes without some sort of
consensus.
Best
Gordon
> > From: Erich Neuwirth
> >
> > I noticed the following:
> > the 2 datasets
> > rep(c(0,1,2),c(10,20,40)) and
> > rep(c(0,1,2),c(10,40,20))
> > produce identical boxplots despite the fact that the medians are
> > different. The reason is that the median in one case
> > coincides with the
> > first quartile, and in the second case with the third quartile.
> > Is there a recommended way of displaying the median visibly in these
> > cases? Setting notch=TRUE displays the median, but does look strange.
---
Dr Gordon K Smyth, Senior Research Scientist, Bioinformatics,
Walter and Eliza Hall Institute of Medical Research,
1G Royal Parade, Parkville, Vic 3050, Australia
Tel: (03) 9345 2326, Fax (03) 9347 0852,
Email: [EMAIL PROTECTED], www: http://www.statsci.org
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] R 2.1.1: read.table processes C-style escapes (PR#8037)
>To: [EMAIL PROTECTED]
>From: Gordon Smyth <[EMAIL PROTECTED]>
>Subject: R 2.1.1: read.table processes C-style escapes
>Date: Wed, 27 Jul 2005 12:51:45 +1000
>
>In R 2.1.1, the default behaviour of scan() was changed to process all
>C-style escapes, even when a delimiter was specified using the 'sep'
>argument. A new argument 'allowEscapes' was introduced to turn this
>processing off.
>
>Because read.table() calls scan(), read.table() inherits the new default
>behaviour of scan() but without a way to turn it off. For example, reading
>a file testdata.txt' containing
>
>X
>A
>\0
>C
>
>produces
>
> > read.delim("testdata.txt")
>[1] X
><0 rows> (or 0-length row.names)
>
>It seems that all the occurances of scan() within read.table() need to
>have 'allowEscapes=FALSE' added to the argument string.
Or, alternatively and perhaps better, scan() could regain some of its
earlier behaviour, to process C-style escapes by default only when 'sep' is
NULL or empty. It seems, to me at least, that C-style escape sequences make
sense only in some sort of source code, and delimited text can't be source
code.
Gordon
>Gordon
>
> > version
> _
>platform i386-pc-mingw32
>arch i386
>os mingw32
>system i386, mingw32
>status Patched
>major2
>minor1.1
>year 2005
>month07
>day 22
>language R
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] R 2.1.1: read.table processes C-style escapes (PR#8037)
Thanks for the reply. Apologies for checking only R patched rather than R-devel. I guess that this means that someone must have a use for allowEscapes=TRUE when read reading a file in table format into a data frame. It is hard to imagine. Gordon At 04:59 PM 27/07/2005, Prof Brian Ripley wrote: >This seems of historical interest only. > >allowEscapes is currently (in R-devel, where development happens) an >argument to read.table. > >We do ask people to check the current version! __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] R 2.1.1: read.table processes C-style escapes (PR#8037)
At 05:21 PM 27/07/2005, Prof Brian Ripley wrote: >On Wed, 27 Jul 2005, Gordon Smyth wrote: > >>Thanks for the reply. Apologies for checking only R patched rather than >>R-devel. >> >>I guess that this means that someone must have a use for >>allowEscapes=TRUE when read reading a file in table format into a data >>frame. It is hard to imagine. > >Octal and hex escapes are a common way to cope with different encodings. Thanks. >I would not want R limited by the imagination of a single user, let alone >a Windows-only user. Oct and hex escapes are not the only uses of backslashes in the computing world, and not only in Windows. TeX markup is one example. In the bioinformatics world, a great variety of ways are used to represent gene annotation information in text files, some of them including backlashes. I find that people find ways to throw all sorts of things at my code. If 'allowEscapes=FALSE' was the default in read.table, then all legacy code using read.table() would continue to behave as in the past. With 'allowEscapes=TRUE', all code ever written which uses read.table(), and which has to cope with backslashes which are not octal or hex escapes, will now be broken. This is not in any sense a complaint, nor a wish that R should remain static or be limited by my imagination. Just a plea that, other things being equal, not breaking legacy code will be a consideration. Gordon __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] fuzzy c-means algorithm in e1071
I was wondering if it was possible to add an extra option to the cmeans function please? In a recent paper by Moore et al.,(2009) they state: "The fuzzy membership function is: 1 - F(Z2), where Z2 is a squared Mahalanobis distance and F is a cumulative chi-square distribution. The Mahalanobis distance is the multivariate equivalent of the standardised random variable Z = (X - M)/S, which is the distance of the univariate random variable X from its mean M normalised by the standard deviation S. In other words, the Mahalanobis distance is a weighted form of the Euclidean, and is preferable because it incorporates the shape of the the distribution of points around the cluster centre (i.e., the geometric shape of the point cloud expressed in terms of variance). The fuzzy membership ranges from 0 to 1 ..." so as well as having a dist = "euclidean" option, would it be possible to have a dist = "mahalanobis" option please? I realise that this is quite an ask - you could even point me in the right direction of where to start editing code and getting it compiled on my own machine. Tim Smyth ref: T.S. Moore et al. Remote Sensing of Environment 113 (2009) 2424-2430 -- View this message in context: http://r.789695.n4.nabble.com/fuzzy-c-means-algorithm-in-e1071-tp3669374p3669374.html Sent from the R devel mailing list archive at Nabble.com. __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] read.table with more cols than headers
I am trying to understand the behaviour of read.table() reading
delimited files (with header=TRUE and fill=TRUE) when there are more
(possibly spurious) columns than headings. I give below four small
data files, all of which have one or two extra columns added to one
line. Reading the first file produces an error message, the second
produces a column of NA, the third adds an extra row, the fourth
ignores the extra columns with no message and no NA. Most
unintuitive! Here are my attempts to understand this, with questions
interpolated.
The behaviour on the first file seems self-explanatory. The number
of headings determines the number of columns, and extra data columns
are not allowed. (On the other hand, the help ?read.table says that
the number of columns is determined from the first five rows, which
suggests that the header line is not the only determiner. If
headers, when present, are indeed the only determiner, perhaps this
should be mentioned in the help. Are headers actually equivalent to
specifying the same set of names using the col.names argument?)
For the second file, the first column is being taken as row
names. This agrees with the help which says if "the header line has
one less entry than the number of columns, the first column is taken
to be the row names". OK, perhaps not the ideal solution for this
data file, but clearly documented behaviour.
In the third file, the extra columns are being taken to be a new
row. This seems wrong, because the help says that cases correspond
to lines. There is no suggestion in the documentation that a line of
the file could contain multiple cases. This is the result I have
most trouble with. I guess could prevent this behaviour by flush=TRUE.
File 4 is curious. Here the number of columns has been determined,
using the first 5 rows of the file, to be two. The extra column on
line 6 can't change this, so the first column doesn't become row
names. But in that case, shouldn't the extra column found on line 6
produce an error message, same as for file 1?
Specifying colClasses to be a vector of length more than 2 when
reading file 3 will produce a result similar to file 4, but with a
warning. It is not clear to me why colClasses should have an
influence, since it doesn't change the determination of the number of
columns. Why a warning here, but an error for file 1 and no message
for file 4?
Any comments gratefully received.
Gordon
X,Y
a,2
b,4,,
c,6
X,Y
a,2
b,4,
c,6
X,Y
a,2
b,4
c,6
d,8
e,10,,
f,12
X,Y
a,2
b,4
c,6
d,8
e,10,
f,12
> read.csv("test1.txt")
Error in read.table(file = file, header = header, sep = sep, quote = quote, :
more columns than column names
> read.csv("test2.txt")
X Y
a 2 NA
b 4 NA
c 6 NA
> read.csv("test3.txt")
X Y
1 a 2
2 b 4
3 c 6
4 d 8
5 e 10
6 NA
7 f 12
> read.csv("test4.txt")
X Y
1 a 2
2 b 4
3 c 6
4 d 8
5 e 10
6 f 12
> read.csv("test3.txt",colClasses=c(NA,NA))
X Y
1 a 2
2 b 4
3 c 6
4 d 8
5 e 10
6 NA
7 f 12
> read.csv("test3.txt",colClasses=c(NA,NA,NA,NA))
X Y
1 a 2
2 b 4
3 c 6
4 d 8
5 e 10
6 f 12
Warning message:
cols = 2 != length(data) = 4 in: read.table(file = file, header =
header, sep = sep, quote = quote,
> sessionInfo()
R version 2.4.0 Under development (unstable) (2006-07-25 r38698)
i386-pc-mingw32
locale:
LC_COLLATE=English_Australia.1252;LC_CTYPE=English_Australia.1252;LC_MONETARY=English_Australia.1252;LC_NUMERIC=C;LC_TIME=English_Australia.1252
attached base packages:
[1] "methods" "stats" "graphics" "grDevices"
"utils" "datasets" "base"
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] S4 Classes
Is there a capability that you would like for the package which could be achieved only if the package was transitioned to S4? If so, explain this to the author. If not, why ask them to change? Gordon >[Rd] S4 Classes >Daniel Gerlanc dgerlanc at gmail.com >Thu Aug 10 23:37:15 CEST 2006 > >Hello All, > >I'm trying to convince someone that they should transition a large project >to use S4 instead of S3 classes. Does anyone have any good citations? >Thanks! > >-- Dan Gerlanc > >-- >Daniel Gerlanc >Williams College '07 __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] Citation for R
This is just a note that R would get a lot more citations if the recommended citation was an article in a recognised journal or from a recognised publisher. I use R in work leading to publications often, and I strongly want to give the R core team credit for their work. However I find that I can't persuade my biological collaborators to include the current R citation (below) in their reference lists, because it is not an article in a recognised journal nor from a recognised publisher. I can cite the 1996 paper by Ihaka and Gentleman, and sometimes this what I do, but I'd really like to give credit to the other R core members as well, for example the CRAN people and those involved in the Windows version. I know this is more work for the R team, like everything else, but an article on the story of R since the creation of the core team would be really nice to see. > citation() To cite R in publications use: R Development Core Team (2005). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org. Gordon __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Citation for R
On Tue, June 14, 2005 1:30 am, Ted Harding said: > Is a journal reference necessary? I have seen many articles where > the statistical software (S-Plus, SPSS, SAS, etc.) was "cited" as > the User Manual, usually only available from the supplier of the > software, sometimes with a WWW URL. Such cases provide a precedent > for R, surely. I have also seen cases where the "citation" was > simply the name of the company (with location, version, date etc.) > > As an example which, as software, is closer to home, consider the > following quotation, and the corresponding citation: > > Quotation: > "This is essentially a four-level hierarchical model and >is easily implemented in, say, WinBUGS (Spiegelhalter, >Thomas and Best, 2000)." [p. 13 of Source reference below] > > Citation Reference: > Spiegelhalter, D. J., Thomas, A. and Best, N. G. (2000) > WinBUGS Version 1.3 User Manual. Cambridge: Medical Research > Council Biostatistics Unit. > (Available from www.mrc-bsu.cam.ac.uk/bugs.) > > Source: > David J. Spiegelhalter, Paul Aylin, Nicola G. Best, > Stephen J. W. Evans, Gordon D. Murray (2002). > Commissioned analysis of surgical performance by using > routine data: lessons from the Bristol inquiry. > J. R. Statist. Soc. A (2002) 165, Part 2, pp. 1-31) > > Surely this would do? Does R need more justification than > WinBUGS? Are JRSS citations less canonical then other journals? Yes, JRSSB citations are less canonical that citations in medical biology journals. Citations are treated very seriously in medical biology world, where impact factors and citations counts are quoted in promotion and grant applications, and there is a reluctance to cite non-refereed publications. An article in Nature would probably not include WinBUGS in the reference list but would simply say in the text that computations were done using "WinBUGS software (Medical Research Council Biostatistics Unit, Cambridge, www.mrc-bsu.cam.ac.uk/bugs)". This might be good enough for the WinBUGS people, but it would not get the WinBUGS manual into the citation system. Note also that R does have a User Guide, i.e., while there is plenty of excellent documentation, there is no single document which is a guide to the whole project. Gordon > Best wishes to all, > Ted. > > > > E-Mail: (Ted Harding) <[EMAIL PROTECTED]> > Fax-to-email: +44 (0)870 094 0861 > Date: 13-Jun-05 Time: 16:30:35 > -- XFMail -- __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Citation for R
On Tue, June 14, 2005 12:49 am, Thomas Lumley said: > On Mon, 13 Jun 2005, Gordon K Smyth wrote: > >> This is just a note that R would get a lot more citations if the >> recommended citation was an article in a recognised journal or from a >> recognised publisher. >> > > This is unfortunately true, but R is *not* an article or a book, it is a > piece of software. I don't think I'm the only person who thinks it is > counterproductive in the long run to encourage users to cite an article > that they probably haven't read instead of citing the software they > actually used. > > Jan's suggestion of the Journal of Statistical Software might provide a > solution, since JSS *does* publish software. > > -thomas In the biology world, it is common to publish an article announcing a software project, and to cite that. The referees of the article are expected to try out and comment on the software. This gives the authors credit, and ensures that both the article and the software have been peer refereed, at least to a limited extent. If the issue of formal citation doesn't worry the R core team, I won't worry about it either. I'm currently revising an article going to Nature which will simply say in the text that quantities "were calculated using R statistical software (http://www.r-project.org)". Gordon __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Standardized Pearson residuals (and score tests)
Hi Peter and others,
If it helps, I wrote a small function glm.scoretest() for the statmod
package on CRAN to compute score tests from glm fits. The score test for
adding a covariate, or any set of covariates, can be extracted very neatly
from the standard glm output, although you probably already know that.
Regards
Gordon
-
Professor Gordon K Smyth,
NHMRC Senior Research Fellow,
Bioinformatics Division,
Walter and Eliza Hall Institute of Medical Research,
1G Royal Parade, Parkville, Vic 3052, Australia.
sm...@wehi.edu.au
http://www.wehi.edu.au
http://www.statsci.org/smyth
Date: Tue, 15 Mar 2011 12:17:46 +0100
From: peter dalgaard
To: Brett Presnell
Cc: r-devel@r-project.org
Subject: Re: [Rd] Standardized Pearson residuals
On Mar 15, 2011, at 04:40 , Brett Presnell wrote:
Background: I'm currently teaching an undergrad/grad-service course
from Agresti's "Introduction to Categorical Data Analysis (2nd edn)"
and deviance residuals are not used in the text. For now I'll just
provide the students with a simple function to use, but I prefer to
use R's native capabilities whenever possible.
Incidentally, chisq.test will have a stdres component in 2.13.0 for
much the same reason.
Thank you. That's one more thing I won't have to provide code for
anymore. Coincidentally, Agresti mentioned this to me a week or two
ago as something that he felt was missing, so that's at least two
people who will be happy to see this added.
And of course, I was teaching a course based on Agresti & Franklin:
"Statistics, The Art and Science of Learning from Data", when I realized
that R was missing standardized residuals.
It would also be nice for teaching purposes if glm or summary.glm had a
"pearsonchisq" component and a corresponding extractor function, but I
can imagine that there might be arguments against it that haven't
occured to me. Plus, I doubt that anyone wants to touch glm unless
it's to repair a bug. If I'm wrong about all that though, ...
Hmm, how would that work? If there was one, I'd worry that people would
start subtracting them which is usually not the right thing to do. I do
miss having a test on the residual deviance occasionally (even though it
is only sometimes meaningful), having to fit a saturated model
explicitly can be a bit silly. E.g. in this case (homogeneity of birth
rates):
anova(glm(births~month,poisson,data=bb), test="Chisq")
...
Df Deviance Resid. Df Resid. Dev P(>|Chi|)
NULL 11 225.98
month 11 225.98 0 0.00 < 2.2e-16 ***
anova(glm(births~1,poisson,data=bb), test="Chisq")
...
Df Deviance Resid. Df Resid. Dev P(>|Chi|)
NULL11 225.98
Notice that the latter version gives me the correct deviance but no
p-value.
A better support for generic score tests could be desirable too. I
suspect that this would actually be the Pearson Chi-square in the
interesting cases.
--
Peter Dalgaard
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd@cbs.dk Priv: pda...@gmail.com
__
The information in this email is confidential and intend...{{dropped:4}}
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Standardized Pearson residuals (and score tests)
On Thu, 17 Mar 2011, peter dalgaard wrote:
On Mar 16, 2011, at 23:29 , Gordon K Smyth wrote:
Hi Peter and others,
If it helps, I wrote a small function glm.scoretest() for the statmod
package on CRAN to compute score tests from glm fits. The score test
for adding a covariate, or any set of covariates, can be extracted very
neatly from the standard glm output, although you probably already know
that.
Thanks Gordon,
I'll have a look. It's the kind of think where you _strongly suspect_
that a neat solution exists, but where you can't just write it down
immediately. Looks like your code needs some elaboration to handle
factor terms and more general model reductions, though.
Yes, the glm.scoretest() function is very basic. At the moment it tests
for adding a single covariate at a time, i.e., a 1 df test, or several 1
df tests. If you like, I could add a multiple column version that would
work for factors etc, it would be only more line or so. I figure you'd
want to pull code out of glm.scoretest() rather than call it explicitly
anyway.
Gordon
-pd
Regards
Gordon
-
Professor Gordon K Smyth,
NHMRC Senior Research Fellow,
Bioinformatics Division,
Walter and Eliza Hall Institute of Medical Research,
1G Royal Parade, Parkville, Vic 3052, Australia.
sm...@wehi.edu.au
http://www.wehi.edu.au
http://www.statsci.org/smyth
--
Peter Dalgaard
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd@cbs.dk Priv: pda...@gmail.com
__
The information in this email is confidential and intend...{{dropped:4}}
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] Will there be 2016 issues of The R Journal?
The R Journal home page doesn't make any promises about how frequently the journal will be published. Historically, though, there have been issues in June and December of each year. The June issue has always appeared by this time (6 August) in previous years. Has there been a change in the publication schedule? Are there still plans for a June 2016 issue? Thanks Gordon - Professor Gordon K Smyth, Head, Bioinformatics Division, Walter and Eliza Hall Institute of Medical Research, http://www.statsci.org/smyth __ The information in this email is confidential and intended solely for the addressee. You must not disclose, forward, print or use it without the permission of the sender. __ __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] Changing style for the Sweave vignettes
Date: Thu, 13 Nov 2014 12:09:47 +0100
From: January Weiner
To: r-devel
Subject: [Rd] Changing style for the Sweave vignettes
As a user, I am always annoyed beyond measure that Sweave vignettes
precede the code by a command line prompt. It makes running examples
by simple copying of the commands from the vignette to the console a
pain. I know the idea is that it is clear what is the command, and
what is the output, but I'd rather precede the output with some kind
of marking.
Is there any other solution possible / allowed in vignettes? I would
much prefer to make my vignettes easier to use for people like me.
Kind regards,
j.
There are different types of people, and some find the default Sweave
format easier.
For a beginner, it is best that the code and output should look the same
in the Sweave vignette as it would look on the screen during an actual
session.
Windows users have access to "paste commands only". For them the default
Sweave vignette format is perfectly convenient -- one can easily cut and
paste whole pages straight from pdf into the R session.
Perhaps a "paste commands only" app for Unix would keep everyone happy :)
Gordon
__
The information in this email is confidential and intend...{{dropped:4}}
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] Using a function from splines.c in our package
Dear all,
I'm writing to ask for advice as to best practice. A PhD student working
with me is writing C++ code that we hope to make public as src code in our
Bioconductor package edgeR. He wants to call the function fmm_spline,
which is part of the source code for the stats package
http://svn.r-project.org/R/trunk/src/library/stats/src/splines.c,
from his C++ code. This function isn't one of the entry points for C code
documented in Chapter 6 of the "Writing R extensions" manual.
We haven't figured out a way to call the fmm_spline function directly from
our C++ code. Is there a way that we have missed?
Can we simply copy the fmm_spline function into our C++ code and declare
where we got it from? Should we include the license declaration from the
header of splines.c? Anything else we need to do satisfying copyright and
be good citizens?
Thanks a lot
Gordon
-
Professor Gordon K Smyth,
Bioinformatics Division,
Walter and Eliza Hall Institute of Medical Research,
1G Royal Parade, Parkville, Vic 3052, Australia.
Tel: (03) 9345 2326, Fax (03) 9347 0852,
http://www.statsci.org/smyth
__
The information in this email is confidential and intend...{{dropped:4}}
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Using a function from splines.c in our package
Thanks. Our package has LGPL (>= 2) licence, so we'll copy with header
and proper acknowledgements.
Gordon
On Mon, 9 Jul 2012, Prof Brian Ripley wrote:
On Mon, 9 Jul 2012, Gordon K Smyth wrote:
Dear all,
I'm writing to ask for advice as to best practice. A PhD student working
with me is writing C++ code that we hope to make public as src code in our
Bioconductor package edgeR. He wants to call the function fmm_spline,
which is part of the source code for the stats package
http://svn.r-project.org/R/trunk/src/library/stats/src/splines.c,
from his C++ code. This function isn't one of the entry points for C code
documented in Chapter 6 of the "Writing R extensions" manual.
None of those are in a package.
We haven't figured out a way to call the fmm_spline function directly from
our C++ code. Is there a way that we have missed?
Not a very portable way, but there are some ideas in 'Writing R Extensions'.
I don't think they are enough for a CRAN or BioC package, though.
Can we simply copy the fmm_spline function into our C++ code and declare
where we got it from? Should we include the license declaration from the
header of splines.c? Anything else we need to do satisfying copyright and
be good citizens?
If your package has a compatible licence, you can (and you must include the
whole header when you copy). CRAN's policies say that when you copy code
from elsewhere you must include that code's authors in the Authors field of
your package, and that clearly is good practice.
Thanks a lot
Gordon
-----
Professor Gordon K Smyth,
Bioinformatics Division,
Walter and Eliza Hall Institute of Medical Research,
1G Royal Parade, Parkville, Vic 3052, Australia.
Tel: (03) 9345 2326, Fax (03) 9347 0852,
http://www.statsci.org/smyth
--
Brian D. Ripley, rip...@stats.ox.ac.uk
Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/
University of Oxford, Tel: +44 1865 272861 (self)
1 South Parks Road, +44 1865 272866 (PA)
Oxford OX1 3TG, UKFax: +44 1865 272595
__
The information in this email is confidential and intend...{{dropped:4}}
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] documentation cross references under R 2.10.0dev for Windows
inkS4class{} expands to
\link[=-class]{}. From (B) it then follows that
there must be an \alias{-class} and a \name{}.
Q1. Is that correct? To me it look a bit inconsistent.
No, \name{} is irrelevant for links. It's the filename that matters in the
3rd form.
Q2. Are there more?
Q3. Will there be more?
Q4. What about
\link[=]{}
\link{:]{}
where can be (almost) any string?
The first is what the 2nd form refers to. "Name" there is what is displayed
in the file making the link.
The second is new, as of 2.10.0, and is the fallback if a filename matching
is not found.
Q4. Are (A) and (B) only supposed to be used for linking within a
package, or can it be used to link to "wherever" exist?
They should work anywhere. The difficulty arises if you link to something
that a user doesn't have installed, or if the link is ambiguous.
Q5. It sounds that (C) and (D) should be avoided. Is that correct?
I think good practice is to make sure that the base of the filename (less
.Rd) is also an alias in the file, and also the \name{} of the file. The
system would probably be less confusing if this were forced, but there are
lots of files out there where it's not true.
You want the filename to be an alias because links sometimes go to aliases
and sometimes to filenames; you want the name to match because that's what is
displayed at the top of the page, so people might remember "just go to the
Foo man page".
Q6. What if exist in two packages 'pkgA' and 'pkgB' and I want
to specify that I mean topic of package 'pkgA', cf namespaces
and pkgA::foo()?
If you follow the good practice above, then use \link[pkgA]{topic}. If you
don't follow that practice, you may be out of luck, because R will look for
the filename topic.Rd in pkgA, not \alias{topic}. However, as of 2.10.0, it
will fall back to the latter.
Q7. I the 1st paragraph above it says "(possibly in another package)"
and in the 3rd paragraph above it is mentioned at "The only reason to
use these forms [...] is to force a reference to a package that might
be further down the search path" - is that the answer to Q4? Will
\link{} be *dynamically* linked to whatever comes first on the
search() path - to reflect the running environment rather than the
intention of the document?
In 2.10.0, I believe this will be the case (but I'd have to check the code to
be sure). I'd recommend being explicit if you are worried about this
possibility.
Reading between the lines, the development of Rd looks exiting.
Instead of 2nd guessing where we are heading, could someone in charge
please give some thoughts on what the plans are and an estimate on how
long it will take before we are there - what R version?
I don't think there's really someone "in charge", but I've been closely
involved with this, so I'll give some thoughts.
Generally speaking, we have releases on a regular schedule, we don't hold
them up for particular features. So I don't think it would be possible to
figure out when development on Rd files will be done. It depends on when
people have the time to do what they think needs doing, and to a large
extent, that depends on how things get used.
Some things that are not there now, but which might be there in the future
(i.e. later than 2.10.0):
- Better support for the \Sexpr macros (which let the content of man pages
depend on R code, executed just before rendering). Right now there's no
special support for that R code; it would make sense to define some functions
to make writing such stuff easier. (This is something that could be done in
a contributed package, it needn't be in base R.)
- Improved prompt() and package.skeleton() functions to take advantage of
the above.
- Graphs in man pages.
- Ways to link from man pages to vignettes. The reverse would be nice, but
it's not possible with the current design, so that would be far off.
- Some general rationalization of the whole help system.
Duncan Murdoch
MISC:
I understand that \link[=-class]{} is part of
standard Rd conventions, but to the best of my knowledge
\link[=.class]{} is not, correct? I would like
to suggest to write a separate paragraph for S4 classes without
mentioning S3 classes. The following also adds to the confusion -
there exists one Rd page with \name{terms} and one with
\name{terms.object}, so it is not really clear what
\link[=terms.object]{terms} is strictly supposed to do - is it of form
\link[=]{} or \link[=]{}. Maybe it is
helpful to clarify what the static/dynamic link will be and what will
be display.
Thanks
Henrik
PS. This is related to today's (Sept 23, 2009) BioConductor posts by
Gordon Smyth - "[Bioc-devel] BioC 2.5: "suspect" interpackage links";
https://stat.ethz.ch/pipermail/bioc-devel/2009-September/001975.html
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] documentation cross references under R 2.10.0dev for Windows
With one expection, all warnings go away when I download the relevant
Bioconductor packages as source code and re-build them (rcmd INSTALL
--build) on my own machine.
The warnings re-appear if I install the Bioconductor packages in the
normal way using biocLite("Biobase") etc. I will follow this up with the
Bioconductor people.
The one exception is the self-reference to limma:00Index. This marked as
a missing link, under Windows only, although it works fine.
Gordon
On Mon, 28 Sep 2009, Gordon K Smyth wrote:
Rcmd check under R 2.10.0dev for Windows seems to be issuing a number of
spurious warning messages about Rd cross-references.
The following warning messages appear when checking the latest (non-public)
version of the Bioconductor package limma. They appear only under Windows,
not Unix or Mac. All the flagged links appear to be ok, in that they
specific a genuine html file, and should therefore not be marked as suspect
or missing.
Regards
Gordon
* using R version 2.10.0 Under development (unstable) (2009-09-27 r49846)
* using session charset: ISO8859-1
* checking Rd cross-references ... WARNING
Missing link(s) in documentation object './man/01Introduction.Rd':
'[limma:00Index]{LIMMA contents page}'
Suspect link(s) in documentation object './man/asmalist.Rd':
'[marray:marrayNorm-class]{marrayNorm}'
Suspect link(s) in documentation object './man/asmatrix.Rd':
'[Biobase]{exprs}'
Suspect link(s) in documentation object './man/dupcor.Rd':
'[statmod]{mixedModel2Fit}'
Suspect link(s) in documentation object './man/EList.Rd':
'[Biobase]{ExpressionSet-class}'
Suspect link(s) in documentation object './man/imageplot.Rd':
'[marray]{maImage}'
Suspect link(s) in documentation object './man/intraspotCorrelation.Rd':
'[statmod]{remlscore}'
Suspect link(s) in documentation object './man/limmaUsersGuide.Rd':
'[Biobase]{openPDF}' '[Biobase]{openVignette}' '[base]{Sys.putenv}'
Suspect link(s) in documentation object './man/malist.Rd':
'[marray:marrayNorm-class]{marrayNorm}'
Suspect link(s) in documentation object './man/normalizebetweenarrays.Rd':
'[marray:maNormScale]{maNormScale}' '[affy:normalize]{normalize}'
Suspect link(s) in documentation object './man/normalizeWithinArrays.Rd':
'[marray:maNorm]{maNorm}'
Suspect link(s) in documentation object './man/normexpfit.Rd':
'[affy:bg.adjust]{bg.parameters}'
Suspect link(s) in documentation object './man/readgal.Rd':
'[marray:read.Galfile]{read.Galfile}'
Suspect link(s) in documentation object './man/rglist.Rd':
'[marray:marrayRaw-class]{marrayRaw}'
On Wed, 23 Sep 2009, Duncan Murdoch wrote:
On 23/09/2009 10:08 PM, Henrik Bengtsson wrote:
Hi,
in 'Writing R Extensions" of R v2.10.0, under Section
'Cross-references' (2009-09-07) it says:
1. "The markup \link{foo} (usually in the combination
\code{\link{foo}}) produces a hyperlink to the help for foo. Here foo
is a topic, that is the argument of \alias markup in another Rd file
(possibly in another package)."
2. "You can specify a link to a different topic than its name by
\link[=dest]{name} which links to topic dest with name name. This can
be used to refer to the documentation for S3/4 classes, for example
\code{"\link[=abc-class]{abc}"} would be a way to refer to the
documentation of an S4 class "abc" defined in your package, and
\code{"\link[=terms.object]{terms}"} to the S3 "terms" class (in
package stats). To make these easy to read, \code{"\linkS4class{abc}"}
expands to the form given above."
3. "There are two other forms of optional argument specified as
\link[pkg]{foo} and \link[pkg:bar]{foo} to link to the package pkg, to
files foo.html and bar.html respectively. These are rarely needed,
perhaps to refer to not-yet-installed packages (but there the HTML
help system will resolve the link at run time) or in the normally
undesirable event that more than one package offers help on a topic20
(in which case the present package has precedence so this is only
needed to refer to other packages). They are only in used in HTML help
(and ignored for hyperlinks in LaTeX conversions of help pages), and
link to the file rather than the topic (since there is no way to know
which topics are in which files in an uninstalled package). The *only*
reason to use these forms for base and recommended packages is to
force a reference to a package that might be further down the search
path. Because they have been frequently misused, as from R 2.10.0 the
HTML help system will look for topic foo in package pkg if it does not
find file foo.html."
Trying to summariz
[Rd] How to capture t-score and p-values from t.test
> Date: Sat, 25 Mar 2006 16:57:50 -0400
> From: Kjetil Brinchmann Halvorsen
> <[EMAIL PROTECTED]>
> Subject: Re: [Rd] How to capture t-score and p-values from t.test
> To: "Bernzweig, Bruce \(Exchange\)" <[EMAIL PROTECTED]>
> Cc: r-devel@r-project.org
>
> Bernzweig, Bruce (Exchange) wrote:
>> When I do t.test on two distributions (see example below), it outputs
>> numerous data about the t.test.
>>
>> What I'd like to do is individually capture some of this data and assign
>> it to other variables.
>>
>> However, I am unable to find anything in the help section.
>
> t.test returns an object of class "htest", but ?htest :
> > ?htest
> No documentation for 'htest' in specified packages and libraries:
> you could try 'help.search("htest")'
> but that does not do anythinh eiather.
>
> Some time ago, I wrote a helpfile for htest, but that was rejected,
> since "S3 classes are not usually documented".
>
> Kjetil
The "htest" object created by t.test() is fully documented in the "Value"
section of
help("t.test"). In effect, this defines what a "htest" object is understood to
be.
help.search("htest") brings up a list of every function which produces an
"htest" object.
Gordon
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
