[R] Estimating bivariate normal density with constrains
Dear R-Users I would like to estimate a constrained bivariate normal density, the constraint being that the means are of equal magnitude but of opposite signs. So I need to estimate four parameters: mu(meanvector (mu,-mu)) sigma_1 and sigma_2 (two sd deviations) rho (correlation coefficient) I have looked at several packages, including Gaussian mixture models in Mclust, but I am not sure what is the best way, or the best package to use for this task. Greatly appreciate any suggestions! Serguei Kaniovski Austrian Institute of Economic Research (WIFO) P.O.Box 91 Tel.: +43-1-7982601-231 1103 Vienna, AustriaFax: +43-1-7989386 Mail: serguei.kaniov...@wifo.ac.at http://www.wifo.ac.at/Serguei.Kaniovski [[alternative HTML version deleted]] __ R-help@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: [R] Accessing variables in a data frame
Thanks! I did not realize you can access variables by name like this. Serguei __ R-help@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.
[R] Accessing variables in a data frame
Hello My data.frame (dat) contains many variables named var.names and others named var.names_var.id For example var.name - c(gdp,inf,unp) var.id - c(w,i) x - paste(var.name, rep(var.id, each=length(var.name)), sep=_) How can I access variables in the dama.frame by names listed in x, for example to compute gdp_w - gdp_i inf_w - inf_i unp_w - unp_i or gdp - gdp_w inf - inf_w unp - unp_w without needing to code each difference separately? Thanks for your help! Serguei Kaniovski [[alternative HTML version deleted]] __ R-help@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.
[R] Canonical link for the GLM gamma model
Hello! When I estimate a glm gamma model using the canonical link (inverse), I get the opposite signs on almost all coefficients compared to the same model (i.e. with the same linear predictor) estimated using other suitable links (log, logit). What confuses me is that most of sources quote the canonical link for the gamma model as 1/mu, but some quote -1/mu! For example: http://support.sas.com/rnd/app/da/new/802ce/stat/chap5/sect19.htm Which is the correct link, and can this explain the sign-flips I get in my models? Thank you very much for your help / advice! Serguei __ R-help@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.
[R] Help on a combinatorial task (lists?)
Hello! I have the following combinatorial problem. Consider the cumulative sums of all permutations of a given weight vector 'w'. I need to know how often weight in a certain position brings the cumulative sums equal or above the given threshold 'q'. In other words, how often each weight is decisive in raising the cumulative sum above 'q'? Here is what I do: w - c(3,2,1) # vector of weights q - 4 # theshold # computes which coordinate of w is decisive in each permutation res - sapply( permn(w), function(x) which(w == x[min(which(cumsum(x) = q))]) ) # complies the frequencies prop.table( tabulate( res )) The problem I have is that when the weights are not unique, the which() function returns a list as opposed to a vector. I don’t know how to proceed when this happens, as tabulate does not work on lists. The answer, of course, should be that equal weights are “decisive” equally often. Can you help? Thanks a lot! Serguei Kaniovski __ R-help@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: [R] Help on a combinatorial task (lists?)
Simple unlist() will not do. In case of repeated weights, unlike permutations of indices permn(1:length(w)) some permutations of weights are identical. E.g. w - c(3,2,2), permutations of indices c(1,2,3) and c(1,3,2) are undistinguishable. I think I have corrected the algorithm, but now I stuck with a rather trivial list manipulation at the very end. library(combinat) w - c(5,3,2,1) i - 1:length(w) q - 7 res - sapply( permn(i), function(x) min(which(cumsum(w[x]) =q)) ) Now I have the vector 'res' ( of size length(permn(i)) ), and I need to extract from each entry of the list produced by permn(i) the element with the index stored in 'res'. E.g. the first three entries: permn(i)[1:3] [[1]] [1] 1 2 3 4 [[2]] [1] 1 2 4 3 [[3]] [1] 1 4 2 3 ... res[1:3] [1] 2 2 3 ... The answer should be 3, 4, 3 ... Thanks again for you help! Serguei K jim holtman schrieb: Does 'unlist' do it for you: w - c(3,3,2,1) # vector of weights q - 4 # theshold # computes which coordinate of w is decisive in each permutation res - unlist(sapply( permn(w), function(x) which(w == x[min(which(cumsum(x) =q))]) )) # complies the frequencies prop.table( tabulate( res )) [1] 0.4 0.4 0.1 0.1 On Tue, Aug 11, 2009 at 7:03 AM, Serguei Kaniovskiserguei.kaniov...@wifo.ac.at wrote: Hello! I have the following combinatorial problem. Consider the cumulative sums of all permutations of a given weight vector 'w'. I need to know how often weight in a certain position brings the cumulative sums equal or above the given threshold 'q'. In other words, how often each weight is decisive in raising the cumulative sum above 'q'? Here is what I do: w - c(3,2,1) # vector of weights q - 4 # theshold # computes which coordinate of w is decisive in each permutation res - sapply( permn(w), function(x) which(w == x[min(which(cumsum(x) = q))]) ) # complies the frequencies prop.table( tabulate( res )) The problem I have is that when the weights are not unique, the which() function returns a list as opposed to a vector. I don’t know how to proceed when this happens, as tabulate does not work on lists. The answer, of course, should be that equal weights are “decisive” equally often. Can you help? Thanks a lot! Serguei Kaniovski __ R-help@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. __ R-help@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.
[R] Problem with contrast
Hello All, I am trying to estimate a generalized linear model using a single dummy variable (bilat). I want to use contr.sum, in which (please correct me if I am wrong) the implicit coefficient on the contrast equals the negative of the sum of all estimated coefficients. I cannot get the contrast to work correctly. The implicit coefficient is way to large (I checked this using the desmat - option dev - package in Stata). I do: contrasts(bilat) - contr.sum(levels(bilat)) mod_1 - gamlss(d ~ bilat, family=BEINF, data=df) but this problem is not specific to gamlss. The same problem occurs also, e.g. in contrasts(bilat) - contr.sum(levels(bilat)) mod_2 - lm(d ~ bilat, data=df) What am I doing wrong? Thank you for help! Serguei Austrian Institute of Economic Research (WIFO) P.O.Box 91 Tel.: +43-1-7982601-231 1103 Vienna, AustriaFax: +43-1-7989386 Mail: serguei.kaniov...@wifo.ac.at http://www.wifo.ac.at/Serguei.Kaniovski [[alternative HTML version deleted]] __ R-help@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: [R] Compute correlation matrix for panel data with specific ordering
I apologize for not being specific enough in my previous posting. Assume you have panel data in the form: df - data.frame( cbind( rep( c( AUT , BEL , DEN , GER ) , 4) , cbind( rep( c( 1999 , 2000 , 2001 , 2002 ) , 4 ) ), sample( 10 , 16 , replace=T) ) ) names(df) - c( country , year , x ) 1. I would like to compute the correlation matrix between countries based on the annual observations of the variable x. I tried the following: library( combinat ) temp - split( df$x, df$year ) apply( combn(4,2) , 2 , function(x) cor( temp[[1]] , temp[[2]] ) ) This gives wrong answer. Why? 2. The pairwise correlations computed as above should be in the order: GER with BEL, GER with DEN, GER with AUT, BEL with DEN, BEL with AUT, DEN with AUT. That is, the correctly sorted vector of factors is: SORT - c( GER , BEL , DEN , AUT ) not c( AUT , BEL , DEN , GER ) May be there is an altogether better way of achieving what I want? Serge __ R-help@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: [R] Compute correlation matrix for panel data with, specific ordering
Ok I see how to sort the factors, but how do I compute the correlation matrix in a repeated observations dataset (see the first part of my question) Thanks again for your help! Serge __ R-help@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.
[R] Compute correlation matrix for panel data with specific ordering
Hello All, I have a panel date - here a small-scale example: df - data.frame(cbind(rep(c(AUT,BEL,DEN,GER),4),cbind(rep(c(1999,2000,2001,2002),4)),sample(10,16,replace=T))) names(df) - c(country,year,x) SORT - c(GER,BEL,DEN,AUT) I need to compute the correlation between countries in the variable x in such a way that the rows columns of the resulting correlation matrix are not in an alphabetical order but in the order of a given factor vector - here SORT. How can I do this? Greatly appreciate any help! Serguei Kaniovski __ R-help@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.
[R] Constrained corr matrix closest to a given corr matrix
Dear All! Is there any code to find a constrained correlation matrix closest in some sense (preferably in the sense of the geometric approximation) to a given correlation matrix? By constrained I mean with some elements constrained, or, simply, set to zero. I have read in a paper that says this can be accomplished with Lagrange multipliers and an optimization routine. Thanks a lot! Serguei Kaniovski __ R-help@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.
[R] Linear constraints for constrasts
Dear List! How can I define contrasts (design matrix) that can all be included, i.e. which do not require a control category be dropped. My application (see below) does not suggest a sensible control category. I am thinking of constraining the (treatment) contrasts to sum up to zero and dropping the constant term in the regression. Is this a good idea? If yes, how to achieve this in R? I am estimating a GLM for bilateral country data. Each observation in on a pair of countries, e.g. GER_USA, GER_JAP, USA_JAP. I constructed the following contrasts: d_GER, d_USA, d_JAP, which take the value of 1 when the country is in the pair and 0 otherwise, i.e. “Bilat”, “d_GER”, “d_USA”, “d_JAP” GER_USA, 1, 1, 0 GER_JAP, 1, 0, 1 USA_JAP, 0, 1, 1 These contrasts highlight the effect of having a given country in the pair. Thank you for your help! Serguei Kaniovski __ R-help@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.
[R] Dummy (factor) based on a pair of variables
Dear All! my data is on pairs of countries, i and j, e.g.: y,i,j 1,AUT,BEL 2,AUT,GER 3,BEL,GER I would like to create a dummy (indicator) variable for use in regression (using factor?), such that it takes the value of 1 if the country is in the pair (i.e. EITHER an i-country OR an j-country). Thank you for your help, Serguei Austrian Institute of Economic Research (WIFO) P.O.Box 91 Tel.: +43-1-7982601-231 1103 Vienna, AustriaFax: +43-1-7989386 Mail: serguei.kaniov...@wifo.ac.at http://www.wifo.ac.at/Serguei.Kaniovski [[alternative HTML version deleted]] __ R-help@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: [R] Dummy (factor) based on a pair of variables
Bernardo: this is not quite what I am looking for, Let the data be: y,i,j 1,AUT,BEL 2,AUT,GER 3,BEL,GER then the dummies sould look like: y,i,j,d_AUT,d_BEL,d_GER 1,AUT,BEL,1,1,0 2,AUT,GER,1,0,1 3,BEL,GER,0,1,1 I can generate the above dummies but can this design be imputed in a reg. model directly? Serguei __ R-help@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.
[R] Implementing a linear restriction in lm()
Dear All! I want to test a coeffcient restriction beta=1 in a univariate model lm (y~x). Entering lm((y-x)~1) does not help since anova test requires the same dependent variable. What is the right way to proceed? Thank you for your help and marry xmas, Serguei Kaniovski Austrian Institute of Economic Research (WIFO) P.O.Box 91 Tel.: +43-1-7982601-231 1103 Vienna, AustriaFax: +43-1-7989386 Mail: serguei.kaniov...@wifo.ac.at http://www.wifo.ac.at/Serguei.Kaniovski [[alternative HTML version deleted]] __ R-help@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: [R] Vertex enumeration and center of mass for convex polytops
Hi Mosche, In my problem the polytope is defined not by its vertices, but by a set of linear equations and inequalities. Serguei Moshe Olshansky schrieb: Hi, If you know that all your points represent vertices of a convex polygon you do not need any special package. The center of mass is just the mean of the coordinates. To enumerate the vertices, compute the vectors from the center of mass to all the vertices. Using atan2 function compute the arguments of all these vectors (between 0 and 2*pi) and number the points according to their argument. --- On Wed, 1/10/08, Serguei Kaniovski [EMAIL PROTECTED] wrote: From: Serguei Kaniovski [EMAIL PROTECTED] Subject: [R] Vertex enumeration and center of mass for convex polytops To: [EMAIL PROTECTED] Received: Wednesday, 1 October, 2008, 10:00 PM Dear All! I am looking for a package that contains routines for vertex enumeration and center of mass computation for convex polytops. Thanks in advance, Serguei Kaniovski __ R-help@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. __ R-help@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.
[R] Vertex enumeration and center of mass for convex polytops
Dear All! I am looking for a package that contains routines for vertex enumeration and center of mass computation for convex polytops. Thanks in advance, Serguei Kaniovski __ R-help@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.
[R] Fitdistr of a 3 parameter Gamma distribution
Hello, how can I fit a three, as opposed to a two parameter Gamma distribution, i.e. one in which shape and rate are not tied. Thanks, Serguei __ R-help@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.
[R] Kappa distribution
Hallo all, I am looking for an R implementation of the four parameter kappa distribution (cdf, pdf, quantile, and ransom numbers), or at a minimum, the generalized logistic distribution. Any suggestions? Thank you very much, Serguei __ R-help@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.
[R] Help with factors
Hallo All, I have difficulties understanding how factors work in R. Suppose a have data in the panel form below. I would to compute a correlation coefficient (actually apply a different function of two time series) in the V variable between members of the two sexes in each city over time. How can this be done? Thank you in advance, Serguei city, year, sex, V 1, 1975, 1, 25.3044 1, 1975, 0, 16.5711 1, 1976, 0, 16.6072 1, 1976, 1, 24.2841 1, 1977, 0, 14.8838 1, 1977, 1, 24.8124 1, 1978, 1, 23.0570 1, 1978, 0, 14.5627 1, 1979, 1, 21.2071 1, 1979, 0, 13.5277 2, 1975, 1, 62.4457 2, 1975, 0, 26.9745 2, 1976, 1, 67.3025 2, 1976, 0, 31.4600 2, 1977, 1, 53.0577 2, 1977, 0, 25.1941 2, 1978, 0, 23.3694 2, 1978, 1, 40.1452 2, 1979, 1, 44.5686 2, 1979, 0, 23.4042 __ R-help@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.
[R] Correlation matrix for data in long format
Hello, I cannot figure out how to use tapply to compute the correlation matrix in the variable x between the states? The data is in long format, e.g.: state,year,x Alabama,2001,0.45 Alabama,2002,0.47 Alabama,2003,0.48 Alabama,2004,0.44 Arizona,2001,0.34 Arizona,2002,0.32 Arizona,2003,0.38 Arizona,2004,0.36 Thank you in advance for your help, Serguei Kaniovski __ R-help@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.
[R] between sum of squares for clusters
Hello, how to compute the between sum of squares the for the clusters obtained by kmeans for the example below? I would like to compute the SSB-SSW based information measures for the clusters, as implemented in cclust, but I am getting an error message. Serguei x1,x2,x3,x4,x5,cl 1.3,0.2,-1.2,3.2,-2.5,1 1.3,-0.4,-1.2,2.8,-1.8,1 -1.4,-0.3,0.7,-0.9,0.5,2 -1,-0.6,1.3,-0.6,0.9,2 -1.4,-0.1,0.8,-0.5,1.1,2 -1,-0.6,0.9,-0.4,1.4,2 0.8,0.9,-1,0.3,-1.1,3 -1.4,-0.8,2.8,-1.1,1.1,4 -0.5,-0.3,-0.1,0.5,-0.3,5 -0.1,1,0,0.1,0.2,5 -0.1,-0.3,-0.4,-0.2,0.4,5 -1.4,2.8,0.7,-0.9,0.6,6 -1,1.8,0.6,-1,1.7,6 0.4,-0.4,0.9,-0.7,0.4,7 0.4,-0.4,0.8,-0.6,0.8,7 0.4,-0.6,0.7,-0.2,1,7 -1.4,1.9,0.6,-0.4,-0.2,8 -1.4,1.7,0.3,-0.4,-0.1,8 [[alternative HTML version deleted]] __ R-help@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.
[R] Quick question on kmeans
Hello, when kmeans draws random vectors for the initial centroids, does it then select the clustering that has emerged most frequently? Serguei __ R-help@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.
[R] Permutations of variables in a dataframe
Hallo All, I would like to apply a function to all permutations of variables in a dataframe (except the first). What is the best way to achieve this? I produce the permutations using: nvar - ncol(dat) - 1 perms - as.matrix( expand.grid(rep( list(1:0) , nvar ))[ , nvar:1] ) Thanks in advance Serguei Test-dataframe, comma-delimited: code,wav,w,area,gdp,def,pop,coast,milspend,agr aut,5,10,83.87,26.39,-1.29,8.07,0,0.72,1.81 bel,1,12,30.53,24.87,-0.28,10.29,0.07,1.29,1.09 bul,7,10,110.91,2.14,1.22,8.03,0.35,1.46,10.88 cyp,6,4,9.25,14.65,-3.26,0.7,0.65,2.11,3.2 cze,6,12,78.87,6.88,-4.44,10.26,0,2,3.19 dnk,2,7,43.09,32.75,2.05,5.34,7.31,1.53,1.98 est,6,4,45.23,5.15,0.82,1.38,3.79,1.58,3.91 fin,5,7,338.15,25.5,3.52,5.18,1.25,1.33,2.87 fra,1,29,547.03,23.99,-2.63,61.05,3.43,2.61,2.39 __ R-help@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.
[R] Information criteria for kmeans
Hello, how is, for example, the Schwarz criterion is defined for kmeans? It should be something like: k - 2 vars - 4 nobs - 100 dat - rbind(matrix(rnorm(nobs, sd = 0.3), ncol = vars), matrix(rnorm(nobs, mean = 1, sd = 0.3), ncol = vars)) colnames(dat) - paste(var,1:4) (cl - kmeans(dat, k)) schwarz - sum(cl$withinss)+ vars*k*log(nobs) Thanks for your help, Serguei Austrian Institute of Economic Research (WIFO) P.O.Box 91 Tel.: +43-1-7982601-231 1103 Vienna, AustriaFax: +43-1-7989386 Mail: [EMAIL PROTECTED] http://www.wifo.ac.at/Serguei.Kaniovski [[alternative HTML version deleted]] __ R-help@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.
[R] Inserting a subsequence between values of a vector
Hallo, suppose I have a vector: x - c(1,1,1,2,2,3,3,3,3,3,4) How can I generate a vector/sequence in which a fixed number of zeroes (say 3) is inserted between the consecutive values, so I get 1,1,1,0,0,0,2,2,0,0,0,3,3,3,3,3,0,0,0,4 thanks a lot, Serguei [[alternative HTML version deleted]] __ R-help@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.
[R] Order observations in a dataframe
Dear All, Suppose I have the following dataframe: country;weight;group bul;10;1 cze;12;1 grc;12;1 hun;12;1 prt;12;1 rom14;1 fra;29;2 ita;29;2 gbr;29;2 aut;10;3 bel;12;3 The group variable denotes the id-number of a group of countries. How can I re-label the groups in the descending order of their cumulative weight, which wound be: country;weight;group fra;29;1 ita;29;1 gbr;29;1 bul;10;2 cze;12;2 grc;12;2 hun;12;2 prt;12;2 rom14;2 aut;10;3 bel;12;3 The first group has the largest sum of weights, the second, with the second largest, and so on. Thanks for your help, Serguei Kaniovski Austrian Institute of Economic Research (WIFO) P.O.Box 91 Tel.: +43-1-7982601-231 1103 Vienna, AustriaFax: +43-1-7989386 Mail: [EMAIL PROTECTED] http://www.wifo.ac.at/Serguei.Kaniovski [[alternative HTML version deleted]] __ R-help@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.
[R] Replacing values job
Hallo, I have two vectors of different lengths which contain the same set of values: X -c(2,6,1,7,4,3,5) Y - c(1,1,6,4,6,1,4,1,2,3,6,6,1,2,4,4,5,4,1,7,6,6,4,4,7,1,2) How can I replace the values in Y with the index (!) of the corresponding values in X. So 2 appears in X in the first coordinate, so all 2âs in Y should be replaced by 1, etc. Thank you for your help, Serguei Austrian Institute of Economic Research (WIFO) P.O.Box 91 Tel.: +43-1-7982601-231 1103 Vienna, AustriaFax: +43-1-7989386 Mail: [EMAIL PROTECTED] http://www.wifo.ac.at/Serguei.Kaniovski [[alternative HTML version deleted]] __ R-help@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.
[R] Vectorize a correlation matrix
Hello I can construct a correlation matrix from an (ordered) vector of correlation coefficients as follows: x - c(0.1,0.2,0.3,0.4,0.5) n - length(x) cmat - diag(rep(0.5,n)) cmat[lower.tri(cmat,diag=0)] - x cmat - cmat+t(cmat) But how to do the reverse operation, i.e. produce x from cmat? Thanks for help, Serguei Kaniovski [[alternative HTML version deleted]] __ R-help@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.
[R] Matrix of dummies from a vector
Hallo From a variable x that defines, say, four classes, I would like to define the matrix mat of dummy variables indicating the classes, i.e. x - c(1,1,1,1,2,2,2,3,3,3,4,4) mat - matrix(c(1,0,0,0, 1,0,0,0, 1,0,0,0, 1,0,0,0, 0,1,0,0, 0,1,0,0, 0,1,0,0, 0,0,1,0, 0,0,1,0, 0,0,1,0, 0,0,0,1, 0,0,0,1), ncol=4, byrow=T) Thank you for your help, Serguei [[alternative HTML version deleted]] __ R-help@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.
[R] How to map clusters to a correlation matrix
Dear All, I have several socio-economic and geographic variables for the 27 EU countries. I would to use these data to derive a correlation matrix between groups of countries (for a different application). I thought of using kmeans to cluster the groups, and then calibrate between group correlations using distances between the centroids, and within group correlations using distances in a cluster to the own centroid. To calibrate is to transform a distance to a (positive) correlation coefficient using some suitable function. Positive correlations reflect the strength of common tendencies among the countries. All the above seems crude to me, especially as you have to choice a transformation function for a distance to a correlation coefficient. Are there any better methods to do this? Thanks is advance, Serguei [[alternative HTML version deleted]] __ R-help@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: [R] How to map clusters to a correlation matrix
To Walter, I am building a model of voting in the EU Council of Ministers. A voting scenario assumes the probabilities of yes votes, possible different for each country, and correlation coefficient between them. The country variables (data) reflect the background characteristics that are relevant for the likelihood of bloc formation in the Council. I would like to cluster the countries in what I call probabilistic voting blocs, and construct a correlation matrix between votes in different blocs and between votes in each bloc. I then use the probabilities and correlation coefficients to construct a joint probability distribution on the set of all conceivable voting outcomes and thus compute the probabilities of voting outcomes that are of interest, such as probability of a country casting a decisive vote, etc. Serguei Austrian Institute of Economic Research (WIFO) P.O.Box 91 Tel.: +43-1-7982601-231 1103 Vienna, AustriaFax: +43-1-7989386 Mail: [EMAIL PROTECTED] http://www.wifo.ac.at/Serguei.Kaniovski [[alternative HTML version deleted]] __ R-help@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.
