Re: [R] confusion matrix - better code?
Dear Monica,
try this:
cm = table(tr, pr)
cm
pr
tr 1 2 3
1 2 1 0
2 2 1 0
3 0 0 3
4 0 1 0
rowSums(cm)
colSums(cm)
Best wishes
Wolfgang Huber
Monica Pisica ha scritto:
Hi,
I�ve written some code to obtain a confusion matrix when the true
classification and the predicted classification are known. Suppose true
classification is called �tr� and predicted classification is �pr�. I have 4
classes in tr, but only 3 classes out of 4 are predicted in �pr�. Following
is my code, but looks quite �clunky� to me. I wonder if you have any
suggestions to improve it.
Thanks,
Monica
-
tr - c(1,2,2,3,3,3,2,4,1,1)
pr-c(1,2,1,3,3,3,1,2,1,2)
dat - data.frame(tr, pr)
class - c(1:length(tr))
m - max(c(length(unique(tr)), length(unique(pr
for(i in 1:length(class)) {
class[i] - sub(' ','',paste(dat[i,1],dat[i,2])) }
dat - data.frame(dat, class)
mat - matrix(0, nrow=m, ncol=m)
for (i in 1:m){
for (j in 1:m){
mat[i,j] - sub(' ','',paste(i,j))
}}
cat - matrix(0, nrow=(m+1), ncol=(m+1))
for (i in 1:m){
for(j in 1:m){
cat[i,j]- nrow(dat[dat$class==mat[i,j],])
}}
for (i in 1:m){
cat[(m+1),i]-sum(cat[1:m,i])
cat[i,(m+1)]- sum(cat[i,1:m])
cat[(m+1),(m+1)] - sum(cat[1:m,(m+1)])
}
cat
[,1] [,2] [,3] [,4] [,5]
[1,]21003
[2,]21003
[3,]00303
[4,]01001
[5,]4330 10
The 5th row / col represents the sum on each row / col respectively.
__
R-help@stat.math.ethz.ch 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: [R] Colours in R, was: deldir package - voronoi
Dear Rolf and Binabina
perhaps this is of use to some:
Colour for Presentation Graphics. Ross Ihaka.
www.stat.auckland.ac.nz/~ihaka/colour/color.pdf
Choosing Color Palettes for Statistical Graphics
Achim Zeileis and Kurt Hornik.
eeyore.ucdavis.edu/stat250/epub-wu-01_abd.pdf
Best wishes
Wolfgang
--
Wolfgang Huber EBI/EMBL Cambridge UK http://www.ebi.ac.uk/huber
Turner ha scritto:
On 12/08/2007, at 1:22 PM, zubin wrote:
Hello!
I am using the deldir package to visualize my data, pretty neat.
However, i need to fill the colors using polycol =, fill with colors
like a heatmap - more of a gradient fill. The only colors i get are
very blocky - how do i assign the correct colors for a gradient,
even a
grayscale? i tried the chart of R colors, using 200 numbers for
grayscale but not getting them. The polycol = colors the cells in the
tesselation with the value a specific vector for color.
snip
I'm sorry, but I can't help here. I've been struggling with colors,
in a different
context, recently myself, and I'm unclear as to how they work. There
are a bunch
of functions --- palette(), colorRamp(), colorRampPalette() that
probably relate
to what you want to do, but I'm not sure just *how* they relate.
With a bit of luck, someone cleverer than I will come to your rescue.
cheers,
Rolf Turner
##
Attention:\ This e-mail message is privileged and confidenti...{{dropped}}
__
R-help@stat.math.ethz.ch 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.
__
R-help@stat.math.ethz.ch 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: [R] sorting with criteria that are out of order
Dear Niels, you can convert the columns (which are apparently character vectors) into ordered factors and then sort / order on these. a [1] b a a a c a a b b c b a c c c c a a a c b=ordered(a, levels=c(b, a, c)) b [1] b a a a c a a b b c b a c c c c a a a c Levels: b a c sort (b) [1] b b b b a a a a a a a a a c c c c c c c Levels: b a c Best wishes Wolfgang -- Wolfgang Huber EBI/EMBL Cambridge UK http://www.ebi.ac.uk/huber Steen Krogh wrote: I try to sort this dataframe: [,1] [,2] [,3][,4] [,5] [,6] [,7] [,8] [1,] CM BARBY INCREASED 0 2 0 1 1 [2,] CM BARBY REDUCED 0 1 2 2 5 [3,] CM BARBY STANDARD 93 51 56 41 77 [4,] CM BONBON INCREASED 43 30 39 32 58 [5,] CM BONBON REDUCED 4 3 6 4 10 [6,] CM BONBON STANDARD 200 141 127 73 134 [7,] RAR BARBY INCREASED 4 1 3 1 5 [8,] RAR BARBY REDUCED 5 7 8 9 16 [9,] RAR BARBY STANDARD 571 286 314 270 467 [10,] RAR BONBON INCREASED 49 92 108 154 240 [11,] RAR BONBON REDUCED11 9 5 6 18 [12,] RAR BONBON STANDARD 978 627 571 324 541 I want the sorting criteria: Column 1: CM before RAR Column 2: BONBON before BARBY Column 3: REDUCED before STANDARD before INCREASED Have played with the order function but without being able to sort out how to sort using information in column three. /Niels Niels Steen Krogh Konsulent ZiteLab ApS Mail: -- [EMAIL PROTECTED] Telefon: --- +45 38 88 86 13 Mobil: - +45 22 67 37 38 Adresse: --- ZiteLab ApS Solsortvej 44 2000 F. Web: --- www.zitelab.dk ZiteLab -Let's Empower Your Data with Webservices __ R-help@stat.math.ethz.ch 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: [R] count the # of appearances...
Dear Bunny, ? table might be what you wish. Best wishes Wolfgang Huber EBI bunny , lautloscrew.com wrote: Hi there, is there a possibility to count the number of appearances of an element in a vector ? i mean of any given element.. deliver all elements which are exactly xtimes in this vector ? thx in advance !! __ R-help@stat.math.ethz.ch 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: [R] Advice on visual graph packages
Hi Jarrett, would the coercion methods for the graph class, provided by the package of the same name at Bioconductor be useful for doing what you want? This is the same class that also Rgraphviz works on. Try library(graph) example(graphNEL-class) as(gR, matrix) class ? graph class ? graphNEL ? toGXL There is a rich sets of methods for setting and accessing node and edge attributes, and it is straightforward R to convert into any other representation you like. See the vignette Attributes for Graph Objects. I am looking at version = 1.13.6 of the package as I write this, Best wishes -- -- Wolfgang Huber EBI/EMBL Cambridge UK http://www.ebi.ac.uk/huber Hey, all. I'm looking for packages that are good at two things 1) Drawing directed graphs (i.e nodes and edges), both with single and double headed arrows, as well as allowing for differences in line width and solid versus dashed. Note: I've tried Rgraphviz here, but have run into some problems (which seem fixable and I may go with it in the end), and it doesn't satisfy need # 2 (which would be ideal if there is a package that does both). 2) Allowing a user to create a directed graph, and have some text object created that can be reprocessed easily reprocessed into a matrix representation, or other representation of my choosing. I've tried dynamicGraph, but it seems buggy, and continually either crashes, behaves very erratically (nodes disappearing when I modify edges), nor is it clear from the UI how one outputs a new graph, nor how one even accesses many graph attributes. This may be my own ignorance on the latter. Do you have any suggestions? Thanks! -Jarrett __ R-help@stat.math.ethz.ch 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: [R] Near function?
Dear Bart,
hclust might be useful for this as well:
dat = c(1,20,2,21)
hc = hclust(dist(dat))
thresh = 2
ct = cutree(hc, h=thresh)
clusteredNumbers = split(dat, ct)
firstOne = dat[!duplicated(ct)]
clusteredNumbers
$`1`
[1] 1 2
$`2`
[1] 20 21
firstOne
[1] 1 20
Best wishes
Wolfgang
I have an integer which is extracted from a dataframe, which is sorted by
another column of the dataframe.
Now I would like to remove some elements of the integer, which are near to
others by their value. For example: integer: c(1,20,2,21) should be c(1,20).
I tried to write a function, but for some reason, somethings won't work
x - 1:20
near - function(x,th) {
nr - NROW(x)
for (i in 1:(nr-1)){
for (j in (i+1):nr){
if (j nr) break
t=0
if (abs(x[i] - x[j]) th) t = 1
if (t== 1) x - x[-j]
if (t== 1) nr - nr-1
if (t== 1) j - (j-1)
cat ( i,i, j,j,\n)
}}
x
}
near(x,10)
This gives you 1 3 7 13 17 while I was suspecting 1, 20 as the outcome.
If you look at the intermediate results of the cat instruction, you see that,
after he substracted a number, he skipped the next one.
Sorting the integer is not an option, the order is important.
I used an integer from 1:20 as an example, while x - sample((1:20),20) is
maybe a bit more representable for our data, but isn't reproducible for the
output of the function.
Maybe there is already an R-function, which does such thing, or what is wrong
with my coding?
thanks a lot for your time
Bart
[[alternative HTML version deleted]]
__
R-help@stat.math.ethz.ch 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.
--
--
Wolfgang Huber EBI/EMBL Cambridge UK http://www.ebi.ac.uk/huber
__
R-help@stat.math.ethz.ch 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: [R] Error in heatmap()
Dear Yuhong, heatmap deals gracefully with sparse occurences of NA in the matrix, but will fail if whole rows or columns are NA. Try preprocessing your xx as follows: xx = xx[rowSums(!is.na(x))!=0, colSums(!is.na(x))!=0] Best wishes Wolfgang -- Wolfgang Huber EBI/EMBL Cambridge UK http://www.ebi.ac.uk/huber Hi, I run into following error when using heatmap() for data matrix xx. Any help is appreciated? xx contains many NAs. hv - heatmap(data.matrix(xx)) Error in hclustfun(distfun(if (symm) x else t(x))) : NA/NaN/Inf in foreign function call (arg 11) Thanks a lot. Yuhong __ R-help@stat.math.ethz.ch 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.
[R] Advanced course R programming and Bioconductor in Cambridge UK 30.3.+1.4.
Dear R users developers, Seth Falcon and Martin Morgan are teaching Advanced R Programming and Bioconductor 30 March - 1 April 2007 at the Wellcome Trust Genome Campus, Hinxton, Cambridge, UK. This two-day course focuses on programming skills required to develop software for statistical analysis of high-throughput genomic data using R and Bioconductor. Lectures and practical sessions introduce the diversity of packages already available for analysis of data, and present the tools and techniques participants need for effectively implementing their own analyses. Closing date for applications: 7th February 2007 Further information and details of how to apply can be found on our website: www.wellcome.ac.uk/advancedcourses and http://www.wellcome.ac.uk/doc_WTX035299.html Best wishes -- Wolfgang Huber EBI/EMBL Cambridge UK http://www.ebi.ac.uk/huber __ R-help@stat.math.ethz.ch 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.
[R] Re: [Rd] (no subject)
Dear Fang Lai, you have sent your mail to both r-devel and r-help. Please do not do this, but decide for one. Cross-posting just creates unnecessary and unpleasant junk-mail to many people. Furthermore, neither the r-devel nor the r-help mailing lists are intended as replacements to taking a basic statistics course or reading the software manuals - rather, as supplement and last resort. The answers are provided by unpaid voluntary contributors, who appreciate that you yourself also make at least a minimal effort before going off the mailing list. Best wishes Wolfgang - Wolfgang Huber European Bioinformatics Institute European Molecular Biology Laboratory Cambridge CB10 1SD England Phone: +44 1223 494642 Http: www.dkfz.de/abt0840/whuber - quote who=fang lai Dear all, I have several questions regarding fisher.test() in R, and I'd highly appreciate any help with it. I have a group of observations, each having people's income, and an indicator of whether selected in or out a program. I want to test the difference between income of people who are in and out. Because the distribution is far from normal, I decide to use the fisher's exact test, using either mean or rank as statistics. Question 0 is: Can I do this test using fisher.test() in R? If so, My first question is: Does fisher.test() offer an option to choose the statistics? Actually it is not clear from the help to me what statistics it uses. Does it just compare the mean of people in and out of the program? My second question is: when the group is large, I always receive a warning message such as Fisher exact result might not be right when I set hybrid=T. When I set hybrid=F, it does return a result of p-value without warning message. I wonder if this p-value is reliable or not. And, how does it get the approximation of p-value when hybrid=F? Ideally, it should randomly draw, say 1000 times, from the full sets of permutation of assignment, and get an approximate p-value--is this the way it works in fisher.test( ) in R? If not, does it use another test, or some other measure of approximation? My last question is: when the group is small enough, will it calculates the exact probabilities even if hybrid=F? Many thanks, Fang = Lai, Fang PhD candidate University of California, Berkeley Department of Agricultural and Resource Economics 314 Giannini Hall, Berkeley, CA 94720-3310 tel: (510) 643 - 5421(O) (510) 847 - 9811(Cell) fax: (510) 643 - 8911 email: [EMAIL PROTECTED] http://www.are.berkeley.edu/jobmarket/fang.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-devel __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
