similar to: displaying pruned clusters

I'm trying to plot observations against observation dates and getting julian dates along the x-axis: library(date) Week<-as.date(c("05/02/03","05/09/03","05/16/03","05/23/03","05/30/03","06/0 7/03","06/14/03")) Leafminers<-c(0,2,5,10,4,6,5) Diglyphus<-c(0,0,4,5,7,3,1) LeafDig<-cbind(Week,Leafminers,Diglyphus)

I'm resending this bug report with a new example. As seen below, cut.dendrogram gives an error message for some heights, but not for others and with some datasets adn not others. I can't see why. Last time I unwittingly sent my message with HTML formatting. This time I'm travelling and using an e-mail system that I am unfamiliar with. As far as I can see, I am not using HTML.

Is there a way of extracting mean distance or dissimilarity for a given strata from a 'dist' or 'dissimilarity' object, e.g. extract mean distances for each species in Anderson's iris data? data(iris) iris.dist<-dist(iris[,1:4]) then what? Mikkel Grum, PhD Genetic Diversity Scientist International Plant Genetic Resources Institute (IPGRI) Sub-Saharan Africa Group ***

Is there a way of removing all rows with missing values from a data frame? I usually use subset(x, var1!="NA") and repeat for each variable. It would be nice to be able to do it in one fell swoop. Also, surprisingly, it doesn't always work. Sometimes I'm left with an empty set even though not all rows have missing values for the variable. Cheers, mikkel Mikkel Grum,

This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_000_01C35A53.75780090 Content-Type: text/plain; charset="iso-8859-1" Anyone have a clue why hclust() and agnes() produce different results in the example below when both use method="average"?? I'm not able to reproduce

I'm trying to plot four different sorghum types on a plot using a different colour/symbol combination for each sorghum type (TYPBOTA in the script below). What am I doing wrong?: plot(xx$LATITUDE,xx$SFD1, points(xx$SFD1,xx$LATITUDE, col=1:4[codes(xx$TYPBOTA)],pch=c(4,3,1,2)[codes(xx$TYPBOTA)]) grid() I get the following error message Error in 1:4[codes(xx$TYPBOTA)] : NA/NaN argument

I have two problems where I've come up with some code that will do the analysis that I want, but it looks pretty clumsy. In the first case, I calculate the variance on five different columns for each of 14 clusters and get them into one matrix. I get the job done, but I would have thought that it could be done in one or two lines, not six, and be generalized so that it didn't matter how

Can anyone tell me what's wrong with this command?? xx<-subset(x,LOCAL.NAME==c("Chifumbata","Chikota"),select=c(13,16,19,23,26,2 9,30,33:48)) Warning message: longer object length is not a multiple of shorter object length in: LOCAL.NAME == c("Chifumbata", "Chikota") both of the following commands work fine

First question: I have an Excel file with both character and numeric variables that I want to import to a data.frame, so I've saved the data to a tab-delimited text file. read.table : * Won't identify the mode of my character variables, as some of them contain only numbers and none of them are surrounded by quotes, * Won't allow me to specify the mode of my variables, * Doesn't

When playing around with reassociate I noticed a seemingly obvious optimization that was not getting done anywhere in llvm… nor in gcc or ICC. Consider the following trivial function: void foo(int a, int b, int c, int d, int e, int *res) { res[0] = (e * a) * d; res[1] = (e * b) * d; res[2] = (e * c) * d; } This function can be optimized down to 4 multiplies instead of 6 by reassociating

I have a specific question and a general question. Specific Question: I want to do an analysis on a data frame by 2 or more class variables (i.e., use 2 or more columns in a dataframe to do statistical classing). Coming from SAS, I'm used to being able to take a data set and have the output of the analysis in a dataset for further manipulation. I have a data set with vote totals, with one

Dear R-community, I was using SVM regression (svm {e1071}) for predictions of single soil properties of a huge data set (3000 samples). There are for the eps-regression using the radial basis kernel three optimization parameters needed. To make things easier (using only two optimization parameters and not loosing performance) I wanted to use LS SVM regression (lssvm{kernlab}). But it

When I use the function cutree(), the numbers of the clusters are not in the same order as the plotted dendrogram. Is there any way of sorting them so that they match the tree? Is it possible to plot a dendrogram horizontally, preferably with the branches to the right? This would enable some practical composite plots, e.g. labels that were an entire table with several columns of information, or

Dear R-programmer, I wrote an adapted implementation of the Kennard-Stone algorithm for sample selection of multivariate data (R 2.7.1 under MacBook Pro, Processor 2.2 GHz Intel Core 2 Duo, Memory 2 GB 667 MHZ DDR2 SDRAM). I used for the heart of the script three embedded loops. This makes it especially for huge datasets very slow. For a datamatrix of 1853*1853 and the selection of 556

Hi! I am trying to calculate a dissimilarity matrix using daisy. The matrix vectver is binary as i test with: > levels(as.factor(vectver)) [1] "0" "1" But the call to daisy gives me the following error message.: > dfl1 <- daisy(vectver, type = list(asymm = c(1:length(vectver[,1])))) Error in daisy(vectver, type = list(asymm = c(1:length(vectver[, 1])))) : at least

Hello, I was using the functino kgs of maptree and I've realized maptree has been removed from CRAN repository. Anybody knows why has it been removed? or where can I find the function kgs that computes the Kelley-Gardner-Sutcliffe penalty function for a hierarchical cluster tree Thank you. [[alternative HTML version deleted]]

I am currently using the function daisy in package cluster to create a dissimilarity matrix because my multivariate dataset contain missing data and variables of various types including factors, symmetric and asymmetric binary and quantitative. This is a step prior to using pco within ecodist. There is a warning which comes twice ">NAs introduced by coercion" I've used

Dear Rhelp,   I did a daisy on 5 lifestyle variables, 3 of which were nominal and 2 were ordinal and assigned types “nominal” and “ordinal” for the variables, respectively.  I got an output indicating their types as “I” for interval(?). Doing it on the Rdata example “flower” gave the same types in the output as the types they were assigned to.  Why is this so? Below are the codes and outputs.  

Hi, I have explored many of the R packages that construct Bayesian trees including the tgp, bart, BMA and maptree packages. I have also searched through some other packages but they do not seem to be suitable for the type of analysis I need to do. I need to construct Bayesian CART that have terminal nodes which have bivariate regressions (not multiple regressions like most of the packages do).

Look at help(sink). I have figured out how to fix up draw.tree to do factor labels correctly and will be doing that in the next few days. Meanwhile I discovered that rpart() provides factor labels in a much more satisfactory way than tree() does. Because of this and because of the additional information rpart provides about the tree building, I recommend you switch to that. > I have created