[R] Different type of legend?
How would I create a legend that looks like the attached image? Basically all of the color boxes are right next to each other and the text is below. This kind of arrangement allows for many more items in the legend. Using the legend() method seems to top out at about 14 items (that will fit in the horizontal plot). Suggestions? Thank you. Kevin __ 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] Different type of legend?
Sorry. I am not sure how to post a link. Basically the legend looks like: * * * * * * * * * * * -4 -3 -2 -1 0 1 2 3 4 Where ' * ' are colored boxes that are right next to each other. Kind of like a gradient. On Mon, Jan 30, 2012 at 3:29 PM, R. Michael Weylandt wrote: Server stripped the attachment. Can you post a link somewhere? Michael On Mon, Jan 30, 2012 at 4:25 PM, rkevinbur...@charter.net wrote: How would I create a legend that looks like the attached image? Basically all of the color boxes are right next to each other and the text is below. This kind of arrangement allows for many more items in the legend. Using the legend() method seems to top out at about 14 items (that will fit in the horizontal plot). Suggestions? Thank you. Kevin __ 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] apply on rows and columns?
I have the following scenario: m - matrix(1:4, ncol=2) m [,1] [,2] [1,]13 [2,]24 apply(m, 2, sum) [1] 3 7 apply(m, 1, sum) [1] 4 6 So I can apply to rows *or* columns. According to the documentation (?apply) MARGIN a vector giving the subscripts which the function will be applied over. E.g., for a matrix 1 indicates rows, 2 indicates columns, c(1, 2) indicates rows and columns. Where X has named dimnames, it can be a character vector selecting dimension names. But I get the following results: apply(m, c(1,2), sum) [,1] [,2] [1,]13 [2,]24 How am I to interpret this result? Thank you. Kevin [[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] List of lists to data frame?
I would like to make the following faster: df - NULL for(i in 1:length(s)) { df - rbind(df, cbind(names(s[i]), time(s[[i]]$series), as.vector(s[[i]]$series), s[[i]]$category)) } names(df) - c(name, time, value, category) return(df) The s object is a list of lists. It is constructed like: s[[object]] - list(. . . . . .) where object would be the name associated with this list s[[i]]$series is a 'ts' object and s[[i]]$category is a name. Constructing this list is reasonably fast but to do some more processing on the data it would be easier if it were converted to a data frame. Right now the above code is unacceptably slow at converting this list of lists to a data frame. Any suggestions on how to optimize this are welcome. Thank you. Kevin [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Formula variable help
I have an R script with the following applicable lines: xshort - window(s, start=st, end=ed) . . . xshort - ts(xshort, frequency=1, start=1) . . . m1 - m2 - m3 - m4 - m5 - m6 - NULL m1 - tslm(xshort ~ trend) I get an error: Error in get(dataname) : object 'xshort' not found When I do traceback() I get: 3: get(dataname) 2: tslm(xshort ~ trend) at #19 1: model.cross.validation(l[[MEN]]$series) Which points to the call to tslm above. Since I am not supply 'data' to the tslm call (in the forecast package),, I am assuming that the code is dying here (in tslm): if (missing(data)) { dataname - as.character(formula)[2] x - get(dataname) data - data.frame(x) colnames(data) - dataname } Any ideas what is failing? Kevin [[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] window?
Thank you. I am glad I asked. It wasn't giving answers that I expected and now I know why. Rather than pull in another package just for this functionality, I will just reassign the frequency by generating a new time series like: dswin - window(ds, start=..., end=...) dswin - ts(dswin, frequency=1) That should work shouldn't it? Thanks again. Kevin On Tue, Nov 8, 2011 at 9:40 PM, R. Michael Weylandt wrote: Like Denis said, you are asking ts to do things that don't make sense; in particular, some of your statements suggest you don't fully understand what window does or what its frequency argument does. Specifically, when you set frequency = 1 in window, that doesn't mean take a window and treat it as if it has frequency 1; rather take the subseries corresponding to yearly observations. Since 53 is prime, there is no regular subseries you can extract with window() other than the original series and the yearly series. ts objects are required to have a frequency so statements like now that I am taking a subset there is no frequency don't really make sense. Take a look at these examples: ## Create some working data ds.53 - ts(rnorm(53*2), frequency=53, start=c(2000,10)) ds.48 - ts(rnorm(48*2), frequency = 48, start = c(2000,10)) ## These all work window(ds.53, frequency = 1) # Returns elements 1 54 of ds.53 window(ds.53, frequency = 53) # Returns every element of ds.53 window(ds.48, frequency = 1) # Returns elements 1 54 of ds.53 window(ds.48, frequency = 12) # Returns elements seq(1, 48, by = 4) of ds.48 window(ds.48, frequency = 48) # Returns every element of ds.48 ## These don't window(ds.53, frequency = 7) window(ds.48, frequency = 9) Here's how you could do the same with xts. library(xts) library(forecast) x = xts(rnorm(53*2), Sys.Date() + 365*seq(0, 2, by = 1/53)) ets(x) # Auto-conversion to ts Michael On Tue, Nov 8, 2011 at 8:19 PM, Kevin Burton rkevinbur...@charter.net wrote: The problem is when I use the window function an try to extract a subset of the time series an specify the frequency as 1 (not only will ets not take a time series with a frequency greater than 24, now that I am taking a subset there is no frequency so I would like to set it to 1 (which is one of the arguments to the window function) but it does not produce what I expect. That is the problem. I fail to see the relationship of the discussion of what frequency is and how to use the forecast package with this problem. -Original Message- From: Dennis Murphy [mailto:djmu...@gmail.com] Sent: Tuesday, November 08, 2011 6:20 PM To: Kevin Burton Cc: R. Michael Weylandt; r-help@r-project.org Subject: Re: [R] window? The ets() function in the forecast package requires either a numeric vector or a Time-Series object (produced from ts()). The frequency argument in ts() refers to the time duration between observations; e.g., frequency = 7 means that the data are weekly; frequency = 12 means that the data are monthly; frequency = 4 means that the data are quarterly. You can see this from the examples on the help page of ts: ?ts at the R prompt. The example associated with the forecast::ets() function uses the USAccDeaths data: data(USAccDeaths) USAccDeaths ## monthly data for six years # Simulate the same structure with ts: u - ts(rnorm(72), start = c(1973, 1), frequency = 12) u # Evidently you want to produce a multivariate series; # here's one way with monthly frequency: v - ts(matrix(rnorm(106), ncol = 2), start = c(2001, 1), frequency = 12) v Is that more or less what you were after? Dennis On Tue, Nov 8, 2011 at 2:04 PM, Kevin Burton rkevinbur...@charter.net wrote: I expect the frequency to be set to what I set it at and the window to return all of the data in the window from the original time series. The error is not because it is prime. I can generate a time series with just 52 values (or 10) and it still occurs. I am building these objects for use with the 'forecast' packages and one of the methods 'ets' cannot handle a frequency above 24 so I set it (or try to) to 1. Will 'window' take z zoo or xts object? Can I convert from zoo or xts to ts? -Original Message- From: R. Michael Weylandt [mailto:michael.weyla...@gmail.com] Sent: Tuesday, November 08, 2011 2:28 PM To: Kevin Burton Cc: r-help@r-project.org Subject: Re: [R] window? I'm not entirely sure that your request makes sense: what do you expect the frequency to be? It makes sense to me as is...Might your troubles be because 53 is prime? More generally, most people don't like working with the raw ts class and prefer the zoo or xts packages because they are much more pleasant for most time series work. You might want to take a look into those. Michael On Tue, Nov 8, 2011 at 3:18 PM, Kevin Burton rkevinbur...@charter.net wrote: This doesn't seem to work: d - rnorm(2*53) ds - ts(d, frequency=53, start=c(2000,10)) dswin - window(ds, start=c(2001,1), end=c(2001,10), frequency=1) dswin
[R] Stack trace?
Currently I have a for loop executing functions and at the end I get a message like: There were 50 or more warnings (use warnings() to see the first 50) If I do what it says and type warnings(), I get 50 messages like: 2: In !is.na(x) !is.na(rowSums(xreg)) : longer object length is not a multiple of shorter object length I am not sure what function these errors are originating from. I don't think it is from any of the 'R' script that I wrote. I would like to see which function is being called when this error is thrown and which called that . . . and so on. I have the same problem with error messages. An error is thrown but I don't have a call stack to help trace down the problem. Is there some function or technique that I could use to help get a call stack? Thank you. Kevin [[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] Serialization help.
I have the following code: c - file(c:/temp/r/SkuSalesModel.br, rb) s - unserialize(c) close(c) rm(c) And it worked as late as yesterday. Today when I came in I get the following error: Error in .Call(R_unserialize, connection, refhook, PACKAGE = base) : negative length vectors are not allowed I have not upgraded or changed any installation and the file has not changed. Any ideas on how I can get more info or solve this error? Thank you. Kevin [[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] Plotting text?
I noticed that the text() command adds text to a plot. Is there a way to either make the plot blank or add text to a blank sheet. I would like to plot a page that contains just text, no plot lines, labels, etc. Suggestions? Kevin [[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] F# vs. R
True, porting old C and Fortran code to C# or F# would be a pain and probably riddled with errors but it is not too soon to start looking to see if there is a better way. There have been numerous ports of LAPACK, BLAS, etc. to C#. Maybe they could be leveraged. Maybe just allowing packages to be wrtten in C# or F# would be helpful. And remember there is Mono. Just my 2 cents. Patrick Burns pbu...@pburns.seanet.com wrote: I'd like to hear answers to this as well. A language doesn't have to be a complete replacement to be useful. F# seems to have some nice features. Pat On 07/07/2010 17:54, Sergey Goriatchev wrote: Hello, Marc No, I do not want to validate Cox PH. :-) I do use R daily, though right now I do not use the statistical part that much. I just generally wonder if any R-user tried F# and his/her opinions. Regards, Sergey On Wed, Jul 7, 2010 at 17:56, Marc Schwartzmarc_schwa...@me.com wrote: On Jul 7, 2010, at 10:31 AM, Sergey Goriatchev wrote: Hello, everyone F# is now public. Compiled code should run faster than R. Anyone has opinion on F# vs. R? Just curious Best, S The key time critical parts of R are written in compiled C and FORTRAN. Of course, if you want to take the time to code and validate a Cox PH or mixed effects model in F# and then run them against R's coxph() or lme()/lmer() functions to test the timing, feel free... :-) So unless there is a pre-existing library of statistical and related functionality for F#, perhaps you need to reconsider your query. Regards, Marc Schwartz -- Patrick Burns pbu...@pburns.seanet.com http://www.burns-stat.com (home of 'Some hints for the R beginner' and 'The R Inferno') __ 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] Forcing scalar multiplication.
I am trying to check the results from an Eigen decomposition and I need to force a scalar multiplication. The fundamental equation is: Ax = lx. Where 'l' is the eigen value and x is the eigen vector corresponding to the eigenvalue. 'R' returns the eigenvalues as a vector (e - eigen(A); e$values). So in order to 'check' the result I would multiply the eigenvalues ('l') by the eigenvectors. But unless I do it one by one (say e$values[1] * e$vectors[,1]) 'R' tries a matrix multiplication and that is not what I want. I would like a matrix that is formed by the SCALAR multiplication of each of the values by the corresponding eigenvector. How can I force such a multiplication? Thank you. Kevin __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Forcing scalar multiplication.
I am trying to check the results from an Eigen decomposition and I need to force a scalar multiplication. The fundamental equation is: Ax = lx. Where 'l' is the eigen value and x is the eigen vector corresponding to the eigenvalue. 'R' returns the eigenvalues as a vector (e - eigen(A); e$values). So in order to 'check' the result I would multiply the eigenvalues ('l') by the eigenvectors. But unless I do it one by one (say e$values[1] * e$vectors[,1]) 'R' tries a matrix multiplication and that is not what I want. I would like a matrix that is formed by the SCALAR multiplication of each of the values by the corresponding eigenvector. How can I force such a multiplication? Thank you. Kevin __ 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] Beginning Eigen System question.
Forgive me if I missunderstand a basic Eigensystem but when I present the following matrix to most any other LinearAlgebra system: 1 3 1 1 2 2 1 1 3 I get an answer like: //$values //[1] 5.00e+00 1.00e+00 -5.536207e-16 //$vectors // [,1] [,2] [,3] //[1,] 0.5773503 -0.8451543 -0.9428090 //[2,] 0.5773503 -0.1690309 0.2357023 //[3,] 0.5773503 0.5070926 0.2357023 But R gives me: //$values //[1] 5.00e+00 1.00e+00 -5.536207e-16 //$vectors // [,1] [,2] [,3] //[1,] -0.5773503 -0.8451543 -0.9428090 //[2,] -0.5773503 -0.1690309 0.2357023 //[3,] -0.5773503 0.5070926 0.2357023 The only difference seems to be the sign on the first eigen vector. What am I missing? Kevin __ 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] Building Hmisc
From the notes I see that for 2.11 Hmisc is not supported and the suggestion is made to build from source. I am on a Windows 7 platform and I got all of the tools and successfully built 'R' from source. I changed to gnuwin32 and entered make all recommended. Even though the tar.gz (source) version of Hmisc has been downloaded and placed in library/recommended folder the build process doesn't seem to pick it up. What do I need to do to build this from source? Right now whenever I start R I get the error: Error: package 'Hmisc' could not be loaded In addition: Warning message: In library(pkg, character.only = TRUE, logical.return = TRUE, lib.loc = lib.loc) : there is no package called 'Hmisc' And I cannot run anything. Thank you for the suggestions. Kevin __ 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] Type-I v/s Type-III Sum-Of-Squares in ANOVA
i believe John Fox offers another solution in his book. Kevin Daniel Wollschlaeger dw...@psychologie.uni-kiel.de wrote: * On Mo, 1. Mar 2010, Ista Zahn wrote: I've posted a short explanation about this at http://yourpsyche.org/miscellaneous that you might find helpful. I'm a As someone who's also struggled with the type X sum of squares topic, I like the idea to completely walk through a numerical example and see what happens. I'd like to extend this a bit, and cover the following aspects: - how are the model comparisons underlying the SS types calculated? - do the compared models obey the marginality principle? - what are the orthogonal projections defining the model comparisons? - are the projections invariant to the type of contrast codes? - are the hypotheses formulated using empirical cell sizes? (are the effect estimates using weighted or unweighted marginal means)? - how can (some of) the SS be calculated without matrix math? Below you'll find the code for SS type III using the 2x2 example from Maxwell and Delaney. For SS type I, II, and III using the 3x3 example in MD, please see http://www.uni-kiel.de/psychologie/dwoll/r/doc/ssTypes.r The model projections for SS type III corresponding to models that violate the marginality principle are not invariant to the coding scheme. If, e.g., Pai is the projection for the model including main effect A and interaction A:B, Pai will be different for non sum-to-zero and sum-to-zero codes. This seems to mean that SS type III for main effects compare different models when using different contrasts codes. Which leads to the question what hypotheses these models actually imply. I'd be grateful if someone could provide any pointers on where to read up on that. I hope this post is not too long! Best, Daniel - # 2x2 unbalanced design: data from Maxwell Delaney 2004 p322 P - 2 # two groups in factor A (female / male) Q - 2 # two groups in factor B (college degree / no degree) g11 - c(24, 26, 25, 24, 27, 24, 27, 23) g12 - c(15, 17, 20, 16) g21 - c(25, 29, 27) g22 - c(19, 18, 21, 20, 21, 22, 19) Y - c(g11, g12, g21, g22) # salary in 100$ A - factor(rep(1:P, c(8+4, 3+7)), labels=c(f, m)) B - factor(rep(rep(1:Q, P), c(8,4, 3,7)), labels=c(deg, noDeg)) - # utility function getInf2x2 (run with different contrasts settings) # fit all relevant regression models for 2x2 between-subjects design # output: * residual sum of squares for each model and their df # * orthogonal projection on subspace as defined # by the design matrix of each model getInf2x2 - function() { X - model.matrix(lm(Y ~ A + B + A:B)) # ANOVA design matrix # indicator variables for factors from design matrix idA - X[ , 2] # factor A idB - X[ , 3] # factor B idI - X[ , 4] # interaction A:B # fit each relevant regression model mod1 - lm(Y ~ 1) # no effect modA - lm(Y ~ idA) # factor A modB - lm(Y ~ idB) # factor B modAB - lm(Y ~ idA + idB) # factors A, B modAI - lm(Y ~ idA + idI) # factor A, interaction A:B modBI - lm(Y ~ idB + idI) # factor B, interaction A:B modABI - lm(Y ~ idA + idB + idI) # full model A, B, A:B # RSS for each regression model from lm() rss1 - sum(residuals(mod1)^2)# no effect, i.e., total SS rssA - sum(residuals(modA)^2)# factor A rssB - sum(residuals(modB)^2)# factor B rssAB - sum(residuals(modAB)^2) # factors A, B rssAI - sum(residuals(modAI)^2) # factor A, A:B rssBI - sum(residuals(modBI)^2) # factor B, A:B rssABI - sum(residuals(modABI)^2) # full model A, B, A:B # degrees of freedom for RSS for each model N - length(Y) # total N df1 - N - (0+1) # no effect:0 predictors + mean dfA - N - (1+1) # factor A: 1 predictor + mean dfB - N - (1+1) # factor B: 1 predictor + mean dfAB - N - (2+1) # factors A, B: 2 predictors + mean dfAI - N - (2+1) # factor A, A:B:2 predictors + mean dfBI - N - (2+1) # factor B, A:B:2 predictors + mean dfABI - N - (3+1) # full model A, B, A:B: 3 predictors + mean --- # alternatively: get RSS for each model and their df manually # based on geometric interpretation # design matrix for each model one - rep(1, nrow(X))# column of 1s X1 - cbind(one) # no effect Xa - cbind(one, idA)# factor A Xb - cbind(one, idB) # factor B Xab - cbind(one, idA, idB) # factors A, B Xai - cbind(one, idA, idI) # factor A, interaction A:B Xbi - cbind(one, idB, idI) # factor B,
[R] Factors attribute?
I noticed that when I fit a linear model using 'lm' there is an attribute called factors that is added to the term. It doesn't seem to appear for 'model.matrix', just 'lm'. I have been unable to find where it gets constructed or what it means? It looks like a two dimensional array that I may be able to use so I would just like to get some 'official' statement regarding what it is and how it is constructed. I would rather not go on my assumptions. An example would be like: l - lm(prestige ~ income + education, data=Duncan) attr(l$terms,factors) income education prestige 0 0 income 1 0 education 0 1 Thank you. Kevin Burton rkevinbur...@charter.net __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Factors attribute?
I am sorry but I didn't see factors mentioned in this documentation. Kevin Henrique Dallazuanna www...@gmail.com wrote: See ?terms On Mon, Mar 22, 2010 at 2:08 PM, rkevinbur...@charter.net wrote: I noticed that when I fit a linear model using 'lm' there is an attribute called factors that is added to the term. It doesn't seem to appear for 'model.matrix', just 'lm'. I have been unable to find where it gets constructed or what it means? It looks like a two dimensional array that I may be able to use so I would just like to get some 'official' statement regarding what it is and how it is constructed. I would rather not go on my assumptions. An example would be like: l - lm(prestige ~ income + education, data=Duncan) attr(l$terms,factors) income education prestige 0 0 income 1 0 education 0 1 Thank you. Kevin Burton rkevinbur...@charter.net __ 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. -- Henrique Dallazuanna Curitiba-Paraná-Brasil 25° 25' 40 S 49° 16' 22 O __ 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] Factors attribute?
I am sorry but I didn't see factors mentioned in this documentation. Kevin Henrique Dallazuanna www...@gmail.com wrote: See ?terms On Mon, Mar 22, 2010 at 2:08 PM, rkevinbur...@charter.net wrote: I noticed that when I fit a linear model using 'lm' there is an attribute called factors that is added to the term. It doesn't seem to appear for 'model.matrix', just 'lm'. I have been unable to find where it gets constructed or what it means? It looks like a two dimensional array that I may be able to use so I would just like to get some 'official' statement regarding what it is and how it is constructed. I would rather not go on my assumptions. An example would be like: l - lm(prestige ~ income + education, data=Duncan) attr(l$terms,factors) income education prestige 0 0 income 1 0 education 0 1 Thank you. Kevin Burton rkevinbur...@charter.net __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Henrique Dallazuanna Curitiba-Paraná-Brasil 25° 25' 40 S 49° 16' 22 O __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Factors attribute format
Thanks to Marc Schultz I found the documentation on the factors attribute under ?term.object. It stats: factors: A matrix of variables by terms showing which variables appear in which terms. The entries are 0 if the variable does not occur in the term, 1 if it does occur and should be coded by contrasts, and 2 if it occurs and should be coded via dummy variables for all levels (as when an intercept or lower-order term is missing). If there are no terms other than an intercept and offsets, this is ‘numeric(0)’. So now this brings up another question. It seems that the attriute is a two dimentional array. When I print it out in 'R' Fitting the formula prestige ~ income + education I get: income education prestige 0 0 income 1 0 education 0 1 This matrix says to me that 'income' occurs in the term 'income' etc. So it seems that this matrix will always be a diagonal matrix with an added row of zeros containing the response term. If the formula is such that the response is a function of one or more of the dependent variables then of course it will be something other that a row of zeros. So far OK? My problem in understanding comes with using a formula that contains R factors. I am using the following (from the TSA package) for an example: l - lm(tempdub ~ season(tempdub)) attr(l$terms, factors) season(tempdub) tempdub 0 season(tempdub) 1 The function 'season' produces a factor (in this case with 12 levels, one for each month). But the factor attribute still has a '1' and not a '2' indicating that the variable should be coded as a dummy variable (factor). Please help my misunderstanding. Thank you. Kevin Burton __ 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] Possible error in anova.lm?
Perhaps those more in the know than I could clarify some confusion. In the ANOVA 'R' code I see: mss - sum(if (is.null(w)) object$fitted.values^2 else w * object$fitted.values^2) if (ssr 1e-10 * mss) warning(ANOVA F-tests on an essentially perfect fit are unreliable) But in the summary.lm I see: mss - if (attr(z$terms, intercept)) sum((f - mean(f))^2) else sum(f^2) or mss - if (attr(z$terms, intercept)) { m - sum(w * f/sum(w)) sum(w * (f - m)^2) } At the very least the ANOVA code seems to be inefficient to calculate mss then only use it for a warning. But it seems that it is being calculated incorrectly. It is hard to tell since it is only used to issue the warning. But it may be that the warning is incorrect in some cases if mss is not calculated correctly. Ideas? Kevin __ 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] Interpretation of 'swtich'
Thatnk you. The documentation indicates as you indicated that if there is not an exact match then the next element is chosen. But it does not indicate the case that contains an exact match but there is not value to be returned (=, case). From what you indicate this is treated as if it was not a match. Kevin Uwe Ligges lig...@statistik.tu-dortmund.de wrote: On 06.03.2010 21:49, rkevinbur...@charter.net wrote: In browsing the source I see the following construct: res- switch(type, working = , response = r, deviance = , pearson = if (is.null(object$weights)) r else r * sqrt(object$weights), partial = r) I understand that 'switch' will execute the code that is matched by its corresponding string value (in this case 'type'). What I don't understand is the empty code. Is this code saying that if the type is deviance then fill the 'res' variable with an empty value? From my naive point of view it seems that 'res' will only get a value(s) if 'type' is 'response', 'pearson', or 'partial'. Please help with my understanding. Please do read the help pages! From ?switch: If there is an exact match then that element is evaluated and returned if there is one, otherwise the next element is chosen, [...] Example: switch(A, A=1, B=, C=2) # 1 switch(B, A=1, B=, C=2) # 2 Uwe Ligges Kevin Burton rkevinbur...@charter.net __ 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] Interpretation of 'swtich'
In browsing the source I see the following construct: res - switch(type, working = , response = r, deviance = , pearson = if (is.null(object$weights)) r else r * sqrt(object$weights), partial = r) I understand that 'switch' will execute the code that is matched by its corresponding string value (in this case 'type'). What I don't understand is the empty code. Is this code saying that if the type is deviance then fill the 'res' variable with an empty value? From my naive point of view it seems that 'res' will only get a value(s) if 'type' is 'response', 'pearson', or 'partial'. Please help with my understanding. Kevin Burton rkevinbur...@charter.net __ 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] What is assign attribute?
I am sorry I still don't understand. In the example you give the 'assign' vecotr is a vector of length 6 and there are indeed 6 columns in the data frame. But the formula only has two variables namely 'Month' and 'Wind'. Where do the values if the 'assign' vector come from? I see '0 1 1 1 1 2'. What is '1' an index to? '2'? Maybe I am get confused with what the term 'factors' does to the formula. Thanks agiain for your help. Kevin Peter Dalgaard p.dalga...@biostat.ku.dk wrote: rkevinbur...@charter.net wrote: I am just curious. Every once and a while I see an attribute attached to an object called assign. What meaning does this have? For example: dist ~ speed, data=cars forms a matrix like: num [1:50, 1:2] 1 1 1 1 1 1 1 1 1 1 ... - attr(*, dimnames)=List of 2 ..$ : chr [1:50] 1 2 3 4 ... ..$ : chr [1:2] (Intercept) speed - attr(*, assign)= int [1:2] 0 1 The dimnames attribute is fairly self-explanatory. I just am not sure what the assign attribute means. It has to do with the mapping between terms of the formula and columns of the design matrix: str(model.matrix(Ozone~factor(Month)+Wind,data=airquality)) num [1:116, 1:6] 1 1 1 1 1 1 1 1 1 1 ... - attr(*, dimnames)=List of 2 ..$ : chr [1:116] 1 2 3 4 ... ..$ : chr [1:6] (Intercept) factor(Month)6 factor(Month)7 factor(Month)8 ... - attr(*, assign)= int [1:6] 0 1 1 1 1 2 - attr(*, contrasts)=List of 1 ..$ factor(Month): chr contr.treatment I.e. columns 2:5 belong to the first non-intercept term, factor(Month). (Notice that it is implicitly assumed that you have the corresponding terms() output to hand, likewise for the contrasts attribute.) -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - (p.dalga...@biostat.ku.dk) FAX: (+45) 35327907 __ 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] Newbie help with ANOVA and lm.
Would someone be so kind as to explain in English what the ANOVA code (anova.lm) is doing? I am having a hard time reconciling what the text books have as a brute force regression and the formula algorithm in 'R'. Specifically I see: p - object$rank if (p 0L) { p1 - 1L:p comp - object$effects[p1] asgn - object$assign[object$qr$pivot][p1] nmeffects - c((Intercept), attr(object$terms, term.labels)) tlabels - nmeffects[1 + unique(asgn)] ss - c(unlist(lapply(split(comp^2, asgn), sum)), ssr) df - c(unlist(lapply(split(asgn, asgn), length)), dfr) } else { ss - ssr df - dfr tlabels - character(0L) } ms - ss/df f - ms/(ssr/dfr) P - pf(f, df, dfr, lower.tail = FALSE) I think I understand the check for 'p' being non-zero. 'p' is essentially the number of terms in the model matrix (including the intercept term if it exists). So in a mathematical description of a regression that included the intercept and one term (like dist ~ speed) you would have a model matrix of a column of '1's and then a column of data. The 'assign' would be a vector containing [0,1]. So then in finding the degrees of freedom you split the asssign matrix with itself. I am having a hard time seeing that this ever produces degrees of freedom that are different. So I get that the vector 'df' would always be something like [2,2,dfr]. But that is obviously wrong. Would someone care to elighten me on what the code above is doing? Thank you. Kevin __ 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] What is assign attribute?
I am just curious. Every once and a while I see an attribute attached to an object called assign. What meaning does this have? For example: dist ~ speed, data=cars forms a matrix like: num [1:50, 1:2] 1 1 1 1 1 1 1 1 1 1 ... - attr(*, dimnames)=List of 2 ..$ : chr [1:50] 1 2 3 4 ... ..$ : chr [1:2] (Intercept) speed - attr(*, assign)= int [1:2] 0 1 The dimnames attribute is fairly self-explanatory. I just am not sure what the assign attribute means. Thank you. Kevin __ 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] What is assign attribute?
I am just curious. Every once and a while I see an attribute attached to an object called assign. What meaning does this have? For example: dist ~ speed, data=cars forms a matrix like: num [1:50, 1:2] 1 1 1 1 1 1 1 1 1 1 ... - attr(*, dimnames)=List of 2 ..$ : chr [1:50] 1 2 3 4 ... ..$ : chr [1:2] (Intercept) speed - attr(*, assign)= int [1:2] 0 1 The dimnames attribute is fairly self-explanatory. I just am not sure what the assign attribute means. Thank you. Kevin __ 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 split.
I read in the documentation for split: ‘split’ divides the data in the vector ‘x’ into the groups defined by ‘f’. But I am still unclear as to its function. Take for example: x - 1:4 split(x, c(0,1)) $`0` [1] 1 3 $`1` [1] 2 4 I am not clear on how this result is reached. Thank you. Kevin __ 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] Building R from source
Thank you for the tip. I was used to inserting write statements and was surpised when it didn't work and reading this section I see that I shouldn't have been doing this anyway. One more question. Is there another call that I can use to print out a 2-dimensional array? Since FORTRAN stores as column major it is hard to print out an array by row as 'R' does. Also it doesn't mention how to not specify the data to be printed out so that just the label is printed. It (the manual) just says it is possible. Thanks again. Kevin Uwe Ligges lig...@statistik.tu-dortmund.de wrote: Since no more information is given: See Writing R Extensions, currently section 6.5: Printing and 6.5.1 Printing from FORTRAN Uwe Ligges On 17.02.2010 03:50, rkevinbur...@charter.net wrote: I found the problem but not a solution. It turns out if I add the following lines to dqrdc2.f I get the error: write(*,300) ldx,n,p 300 format(3i4) I don't get a compile error but I get the seemingly unrelated error in linking R.DLL I guess the question now is, How do I add a simple print statement?. Or, what is wrong with the above print statement? Thank you. Kevin __ 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] qr test?
I am testing 'qr' with an admittedly contrived matrix and I am getting different results than I am from another package. The matrix that I am using is: x - matrix(seq(.1, by=.1, length.out=12), 4) So the whole test is: x - matrix(seq(.1, by=.1, length.out=12), 4) qr(x) And the output from 'R' is: $qr [,1] [,2] [,3] [1,] -0.5477226 -1.2780193 -2.008316e+00 [2,] 0.3651484 -0.3265986 -6.531973e-01 [3,] 0.5477226 -0.3781696 -1.650163e-16 [4,] 0.7302967 -0.9124744 8.078153e-01 $rank [1] 2 $qraux [1] 1.182574 1.156135 1.589436 $pivot [1] 1 2 3 attr(,class) [1] qr The differences that I see is in the last value of qraux. I was expecting 1.83205 not 1.589436. Also the last row of the decomposition shows: [4,] 0.7302967 -0.9124744 8.078153e-01 I was expecting 0.73030-0.91247-0.55470 So again it is the last element of the array. For the linear algebra gurus out there is this to be expected from the contrived matrix? Is there a better matrix that I can use to test that will more or less give consistent agreed upon results for a QR decomposition? Thank you. Kevin __ 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] Problems building from sources
I am trying to build R-2.9.2 from source on a Windows 7 machine. I have installed all of the requisite software and followed the instructions. I also could have sworn that I had a successful build. But now I get the following error. gcc -std=gnu99 -shared -s -mwindows -o R.dll R.def console.o dataentry.o dynload .o edit.o editor.o embeddedR.o extra.o opt.o pager.o preferences.o psignal.o rho me.o rt_complete.o rui.o run.o shext.o sys-win32.o system.o dos_wglob.o e_pow.o malloc.o ../main/libmain.a ../appl/libappl.a ../nmath/libnmath.a getline/gl.a .. /extra/xdr/libxdr.a ../extra/pcre/libpcre.a ../extra/bzip2/libbz2.a ../extra/int l/libintl.a ../extra/trio/libtrio.a ../extra/tzone/libtz.a dllversion.o -L. -lgf ortran -lRblas -L../../bin -lRzlib -lRgraphapp -lRiconv -lcomctl32 -lversion c:/rtools/mingw/bin/../lib/gcc/mingw32/4.2.1-sjlj/../../../libmingwex.a(mingw_sn printf.o):mingw_snprintf.c:(.text+0x1970): multiple definition of `snprintf' ../extra/trio/libtrio.a(compat.o):compat.c:(.text+0x30): first defined here c:/rtools/mingw/bin/../lib/gcc/mingw32/4.2.1-sjlj/../../../libmingwex.a(mingw_sn printf.o):mingw_snprintf.c:(.text+0x170): multiple definition of `vsnprintf' ../extra/trio/libtrio.a(compat.o):compat.c:(.text+0x20): first defined here collect2: ld returned 1 exit status make[3]: *** [R.dll] Error 1 make[2]: *** [../../bin/R.dll] Error 2 make[1]: *** [rbuild] Error 2 make: *** [all] Error 2 It seems like vsnprint and snprintf are multiply defined. Any ideas? Kevin Dimitri Shvorob dimitri.shvo...@gmail.com wrote: Now that we have a reproducible example... ;) -- View this message in context: http://n4.nabble.com/Problems-with-boxplot-in-ggplot2-qplot-tp1555338p1557994.html Sent from the R help mailing list archive at Nabble.com. __ 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] Building R from source
I found the problem but not a solution. It turns out if I add the following lines to dqrdc2.f I get the error: write(*,300) ldx,n,p 300 format(3i4) I don't get a compile error but I get the seemingly unrelated error in linking R.DLL I guess the question now is, How do I add a simple print statement?. Or, what is wrong with the above print statement? Thank you. Kevin __ 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] F77_CALL, F77_NAME definition
I give up. Maybe it is my search (Windows) but I cannot seem to find the definition of the F77_CALL or F77_NAME macros. Either there are too many matches or the search just doesn't find it. For example where is the source for: F77_CALL(dpotri) ? Thank you. Kevin __ 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] write.csv and header
I have a small request regarding this append feature. As it is now if the data is appended to the file so is the header. I would like to have the header only entered once and appends just append the data. Doable? Kevin Patrick Connolly p_conno...@slingshot.co.nz wrote: On Tue, 15-Dec-2009 at 01:55PM +0100, Gustaf Rydevik wrote: | Hi, | | ?write.table and the argument append should be of help. | example: | | sink(test.csv) | cat(-) | cat(\n) | cat(This is \n a test of header) | cat(\n) | cat(-) | cat(\n) | sink() | Or, instead of the use of sink() and cat(), one could use the write() function, also using the append = TRUE argument. | write.table(matrix(rnorm(100),nrow=10),file=test.csv,append=TRUE,sep=,) | | | regards, | Gustaf | | | -- | Gustaf Rydevik, M.Sci. | tel: +46(0)703 051 451 | address:Essingetorget 40,112 66 Stockholm, SE | skype:gustaf_rydevik | |[[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. -- ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~. ___Patrick Connolly {~._.~} Great minds discuss ideas _( Y )_ Average minds discuss events (:_~*~_:) Small minds discuss people (_)-(_). Eleanor Roosevelt ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~. __ 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.
Re: [R] write.csv and header
I have a small request regarding this append feature. As it is now if the data is appended to the file so is the header. I would like to have the header only entered once and appends just append the data. Doable? Kevin Patrick Connolly p_conno...@slingshot.co.nz wrote: On Tue, 15-Dec-2009 at 01:55PM +0100, Gustaf Rydevik wrote: | Hi, | | ?write.table and the argument append should be of help. | example: | | sink(test.csv) | cat(-) | cat(\n) | cat(This is \n a test of header) | cat(\n) | cat(-) | cat(\n) | sink() | Or, instead of the use of sink() and cat(), one could use the write() function, also using the append = TRUE argument. | write.table(matrix(rnorm(100),nrow=10),file=test.csv,append=TRUE,sep=,) | | | regards, | Gustaf | | | -- | Gustaf Rydevik, M.Sci. | tel: +46(0)703 051 451 | address:Essingetorget 40,112 66 Stockholm, SE | skype:gustaf_rydevik | |[[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. -- ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~. ___Patrick Connolly {~._.~} Great minds discuss ideas _( Y )_ Average minds discuss events (:_~*~_:) Small minds discuss people (_)-(_). Eleanor Roosevelt ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~. __ 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] Normal tests disagree?
If I have data that I feed into shapio.test and jarque.bera.test yet they seem to disagree. What do I use for a decision? For my data set I have p.value of 0.05496421 returned from the shapiro.test and 0.882027 returned from the jarque.bera.test. I have included the data set below. Thank you. Kevin Category,Period,Residual CHILD HATS, WIGS MASKS,1/1/2005,-0.449735723758323 CHILD HATS, WIGS MASKS,2/1/2005,0.281461045050074 CHILD HATS, WIGS MASKS,3/1/2005,0.591383050911335 CHILD HATS, WIGS MASKS,4/1/2005,0.239998659520616 CHILD HATS, WIGS MASKS,5/1/2005,0.00343879474063987 CHILD HATS, WIGS MASKS,6/1/2005,-2.64372061292663 CHILD HATS, WIGS MASKS,7/1/2005,0.381630655290173 CHILD HATS, WIGS MASKS,8/1/2005,-1.79543281552347 CHILD HATS, WIGS MASKS,9/1/2005,1.90631012440313 CHILD HATS, WIGS MASKS,10/1/2005,-0.256232543929779 CHILD HATS, WIGS MASKS,11/1/2005,1.83452602676812 CHILD HATS, WIGS MASKS,12/1/2005,-1.06869719416837 CHILD HATS, WIGS MASKS,1/1/2006,1.04378655286183 CHILD HATS, WIGS MASKS,2/1/2006,0.232655831328322 CHILD HATS, WIGS MASKS,3/1/2006,-0.939084802643773 CHILD HATS, WIGS MASKS,4/1/2006,0.854132879285335 CHILD HATS, WIGS MASKS,5/1/2006,-1.71217066877156 CHILD HATS, WIGS MASKS,6/1/2006,1.28040273099582 CHILD HATS, WIGS MASKS,7/1/2006,-0.386415431325857 CHILD HATS, WIGS MASKS,8/1/2006,-0.769127669783483 CHILD HATS, WIGS MASKS,9/1/2006,-0.810996835089867 CHILD HATS, WIGS MASKS,10/1/2006,0.0477292147635991 CHILD HATS, WIGS MASKS,11/1/2006,0.294672848750557 CHILD HATS, WIGS MASKS,12/1/2006,-0.0841330473924862 CHILD HATS, WIGS MASKS,1/1/2007,0.231663729192233 CHILD HATS, WIGS MASKS,2/1/2007,-0.601790650547443 CHILD HATS, WIGS MASKS,3/1/2007,0.285635768516625 CHILD HATS, WIGS MASKS,4/1/2007,-0.963154959558619 CHILD HATS, WIGS MASKS,5/1/2007,1.52188112949994 CHILD HATS, WIGS MASKS,6/1/2007,-0.826092842933196 CHILD HATS, WIGS MASKS,7/1/2007,1.91937201229077 CHILD HATS, WIGS MASKS,8/1/2007,-0.317789483136924 CHILD HATS, WIGS MASKS,9/1/2007,-0.865011007394312 CHILD HATS, WIGS MASKS,10/1/2007,-0.0281604973711276 CHILD HATS, WIGS MASKS,11/1/2007,-0.123887049811822 CHILD HATS, WIGS MASKS,12/1/2007,-0.0327727730592468 CHILD HATS, WIGS MASKS,1/1/2008,-0.0654939600771254 CHILD HATS, WIGS MASKS,2/1/2008,0.279247739913908 CHILD HATS, WIGS MASKS,3/1/2008,0.167606602923418 CHILD HATS, WIGS MASKS,4/1/2008,0.189533097427477 CHILD HATS, WIGS MASKS,5/1/2008,0.402062194225847 CHILD HATS, WIGS MASKS,6/1/2008,1.97150984262995 CHILD HATS, WIGS MASKS,7/1/2008,-2.27538477532968 CHILD HATS, WIGS MASKS,8/1/2008,1.89091792097945 CHILD HATS, WIGS MASKS,9/1/2008,0.0251732151287081 CHILD HATS, WIGS MASKS,10/1/2008,-0.2349741808124 CHILD HATS, WIGS MASKS,11/1/2008,-0.659332058368173 CHILD HATS, WIGS MASKS,12/1/2008,0.127284768034285 CHILD HATS, WIGS MASKS,1/1/2009,-1.42838560676513 CHILD HATS, WIGS MASKS,2/1/2009,0.617689775286461 CHILD HATS, WIGS MASKS,3/1/2009,-0.034243005247084 CHILD HATS, WIGS MASKS,4/1/2009,-0.304574261133836 CHILD HATS, WIGS MASKS,5/1/2009,0.128679369916751 CHILD HATS, WIGS MASKS,6/1/2009,-0.657479389968652 CHILD HATS, WIGS MASKS,7/1/2009,0.608766068692517 CHILD HATS, WIGS MASKS,8/1/2009,1.92814770869400 CHILD HATS, WIGS MASKS,9/1/2009,-0.172644961366165 CHILD HATS, WIGS MASKS,10/1/2009,-0.453255508263169 CHILD HATS, WIGS MASKS,11/1/2009,-1.09903330959344 __ 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] Normal tests disagree?
If I have data that I feed into shapio.test and jarque.bera.test yet they seem to disagree. What do I use for a decision? For my data set I have p.value of 0.05496421 returned from the shapiro.test and 0.882027 returned from the jarque.bera.test. I have included the data set below. Thank you. Kevin Category,Period,Residual CHILD HATS, WIGS MASKS,1/1/2005,-0.449735723758323 CHILD HATS, WIGS MASKS,2/1/2005,0.281461045050074 CHILD HATS, WIGS MASKS,3/1/2005,0.591383050911335 CHILD HATS, WIGS MASKS,4/1/2005,0.239998659520616 CHILD HATS, WIGS MASKS,5/1/2005,0.00343879474063987 CHILD HATS, WIGS MASKS,6/1/2005,-2.64372061292663 CHILD HATS, WIGS MASKS,7/1/2005,0.381630655290173 CHILD HATS, WIGS MASKS,8/1/2005,-1.79543281552347 CHILD HATS, WIGS MASKS,9/1/2005,1.90631012440313 CHILD HATS, WIGS MASKS,10/1/2005,-0.256232543929779 CHILD HATS, WIGS MASKS,11/1/2005,1.83452602676812 CHILD HATS, WIGS MASKS,12/1/2005,-1.06869719416837 CHILD HATS, WIGS MASKS,1/1/2006,1.04378655286183 CHILD HATS, WIGS MASKS,2/1/2006,0.232655831328322 CHILD HATS, WIGS MASKS,3/1/2006,-0.939084802643773 CHILD HATS, WIGS MASKS,4/1/2006,0.854132879285335 CHILD HATS, WIGS MASKS,5/1/2006,-1.71217066877156 CHILD HATS, WIGS MASKS,6/1/2006,1.28040273099582 CHILD HATS, WIGS MASKS,7/1/2006,-0.386415431325857 CHILD HATS, WIGS MASKS,8/1/2006,-0.769127669783483 CHILD HATS, WIGS MASKS,9/1/2006,-0.810996835089867 CHILD HATS, WIGS MASKS,10/1/2006,0.0477292147635991 CHILD HATS, WIGS MASKS,11/1/2006,0.294672848750557 CHILD HATS, WIGS MASKS,12/1/2006,-0.0841330473924862 CHILD HATS, WIGS MASKS,1/1/2007,0.231663729192233 CHILD HATS, WIGS MASKS,2/1/2007,-0.601790650547443 CHILD HATS, WIGS MASKS,3/1/2007,0.285635768516625 CHILD HATS, WIGS MASKS,4/1/2007,-0.963154959558619 CHILD HATS, WIGS MASKS,5/1/2007,1.52188112949994 CHILD HATS, WIGS MASKS,6/1/2007,-0.826092842933196 CHILD HATS, WIGS MASKS,7/1/2007,1.91937201229077 CHILD HATS, WIGS MASKS,8/1/2007,-0.317789483136924 CHILD HATS, WIGS MASKS,9/1/2007,-0.865011007394312 CHILD HATS, WIGS MASKS,10/1/2007,-0.0281604973711276 CHILD HATS, WIGS MASKS,11/1/2007,-0.123887049811822 CHILD HATS, WIGS MASKS,12/1/2007,-0.0327727730592468 CHILD HATS, WIGS MASKS,1/1/2008,-0.0654939600771254 CHILD HATS, WIGS MASKS,2/1/2008,0.279247739913908 CHILD HATS, WIGS MASKS,3/1/2008,0.167606602923418 CHILD HATS, WIGS MASKS,4/1/2008,0.189533097427477 CHILD HATS, WIGS MASKS,5/1/2008,0.402062194225847 CHILD HATS, WIGS MASKS,6/1/2008,1.97150984262995 CHILD HATS, WIGS MASKS,7/1/2008,-2.27538477532968 CHILD HATS, WIGS MASKS,8/1/2008,1.89091792097945 CHILD HATS, WIGS MASKS,9/1/2008,0.0251732151287081 CHILD HATS, WIGS MASKS,10/1/2008,-0.2349741808124 CHILD HATS, WIGS MASKS,11/1/2008,-0.659332058368173 CHILD HATS, WIGS MASKS,12/1/2008,0.127284768034285 CHILD HATS, WIGS MASKS,1/1/2009,-1.42838560676513 CHILD HATS, WIGS MASKS,2/1/2009,0.617689775286461 CHILD HATS, WIGS MASKS,3/1/2009,-0.034243005247084 CHILD HATS, WIGS MASKS,4/1/2009,-0.304574261133836 CHILD HATS, WIGS MASKS,5/1/2009,0.128679369916751 CHILD HATS, WIGS MASKS,6/1/2009,-0.657479389968652 CHILD HATS, WIGS MASKS,7/1/2009,0.608766068692517 CHILD HATS, WIGS MASKS,8/1/2009,1.92814770869400 CHILD HATS, WIGS MASKS,9/1/2009,-0.172644961366165 CHILD HATS, WIGS MASKS,10/1/2009,-0.453255508263169 CHILD HATS, WIGS MASKS,11/1/2009,-1.09903330959344 __ 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] loess smoothing
Hello, In reading the loess description I see: span: the parameter alpha which controls the degree of smoothing. The default seems to be 0.75. Would it be possible to expand on this decription so I can avoid trail and error? Can I increase this pass 'span' 1? Qualitatively to what degree changing this value affects the smoothing of the data? Thank you. Kevin __ 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] uniroot vs.optimize
I looked at the descriptions for uniroot and optimize and they are somewhat different but the book reference is the same and I am wondering if there are reasons to pick one over the other? Thank you. Kevin __ 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] Normal distribution test
This is probably an even more basic question but shapiro.test return both the statistic (w) and the significance (pw) of the statistic. For this test the null-hypothesis is that the distirbution is not normal so very small values of pw would mean that there is very little chance that the distiribution is not normal. Correct? Thank you. Kevin Johannes Graumann johannes_graum...@web.de wrote: Markus Mehrwald wrote: Hi all, I am completely new to R and my knowledge of statistics is quite small so I hope you can help my. I have three dimensional point data which represents (and this is what I do not know for sure) a normal distribution. Now I want to test if this is true or not and as I can remember from statistics lessons I can use Chi-Square test for distribution test. BUT: I have realy no idea how to do this with R and additionally if my assumptions are correct and if this is possible with R at all. Thank you very much in advance for any answer. Markus See ?shapiro.test or ?ks.test HTH, Joh __ 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] Relase positive with log and zero of negative with 0
This is a very simple question but I couldn't form a site search quesry that would return a reasonable result set. Say I have a vector: x - c(0,2,3,4,5,-1,-2) I want to replace all of the values in 'x' with the log of x. Naturally this runs into problems since some of the values are negative or zero. So how can I replace all of the positive elements of x with the log(x) and the rest with zero? Thank you. Kevin __ 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] Building from source under Windows 7
I have downloaded all of the tools and read the readme's that I know about but I am still getting the following error when I try to build from source: C:\Program Files (x86)\R\R-2.9.2\src\gnuwin32make all recommended make[1]: `Rpwd.exe' is up to date. cp -p etc/Makeconf etc/Rcmd_environ etc/Rconsole etc/Rdevga etc/Rprofile.site et c/rgb.txt ../../../etc cp: preserving permissions for `../../../etc/Makeconf': Permission denied cp: preserving permissions for `../../../etc/Rcmd_environ': Permission denied cp: preserving permissions for `../../../etc/Rconsole': Permission denied cp: preserving permissions for `../../../etc/Rdevga': Permission denied cp: preserving permissions for `../../../etc/Rprofile.site': Permission denied cp: preserving permissions for `../../../etc/rgb.txt': Permission denied make[3]: *** [fixetc] Error 1 make[2]: *** [fixfiles] Error 2 make[1]: *** [rbuild] Error 2 make: *** [all] Error 2 C:\Program Files (x86)\R\R-2.9.2\src\gnuwin32 I have given myself full control on all of the directories and files yet I am unable to get rid of this 'Permission denied' error. Anybody out there that has similar problem building 2.9.2 under Windows 7 x64? Thank you. Kevin __ 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] Formula with in memory data.frame
I have an array of data.frame(s) that I would like to smooth with loess one at a time. the array is master and the two variable that I am interested in is Period and Quantity. So my first attempt at calling loess is: loess(Quantity ~ Period, master[[i]]) But I get the following error: Error: NA/NaN/Inf in foreign function call (arg 2) In addition: Warning message: NAs introduced by coercion I did a str on the array element str(master[[j]]): 'data.frame': 58 obs. of 3 variables: Factor w/ 41 levels 10\ Plates,..: 1 1 1 1 1 1 1 1 1 1 ... $ Period : POSIXct, format: 0001-01-20 0002-01-20 ... $ Quantity: int 0 0 0 0 0 0 0 0 0 0 ... So am I doing something wrong? Thank you. Kevin __ 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] Non-normal residuals.
Hello, I asked a question about what the most likely process to follow if after a time-series fit is performed the residuals are found to be non-normal. One peron responded and offered to help if I supplied a sample data set. Unfortunately now that I have a sample I have lost the emai addressl. If you are that person or have some ideas please email me back at rkevinbur...@charter.net. Thank you. Kevin __ 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] Year and Month extraction from Date object.
Hello, I have seen much discussion on Date. But I can't seem to do this simple operation. I can convert a string to a date: d - as.Date(DATE, format=%m/%d/%Y) But what I want to do is extract the year and month so I can construct an element in a ts object. Ideally I would like to see d$year but that doesn't seem to be available. Once I have a Date object how can I get an integer year? Thank you. Kevin __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] PDF Corrupted?
I am running R 2.9.2 and creating a PDF that I am trying to open with Adobe Reader 9.2 but when I try to open it the reader responds with There was an error opening this document. The file is damaged and cannot be repaired.: I am using the R command(s): pdf(file=cat.pdf, title=Historical Sales By Category) for(j in 1:length(master)) { d - as.Date(master[[j]]$Period[1], format=%m/%d/%Y) fit - ets(ts(master[[j]]$Quantity, start=c(1900 + as.POSIXlt(d)$year, 1 + as.POSIXlt(d)$mon), frequency=12)) plot(fit, col.axis = sky blue, col.lab = thistle) title(master[[j]]$Category, cex.main = 2, font.main= 4, col.main= blue) } Any idea what I am doing wrong? Thank you. Kevin __ 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] Time forecasting next step.
This is kind of a general question about methodology more than anything. But I was looking for fome advice. I have fit a time-series model and feel pretty confident that I have taken this model (exponential smoothing) as far as it will go. In other words looking at the data and the fitted curves I think it is as close as I can get. But when I plot the residuals and form a qqplot it seems that the residuals are not normal. From the QQ-plot there is some factor that is influencing the series that cannot be attributed to noramal random fluxuation. I can run 'tsdiag' to determine basically whether the residuals are normall and random, but what if they are not? What would be the next set of 'R' commands that I might run to find this influence? Any suggestions? Kevin __ 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] Tinn-R setup
I recently installed R 2.9.2 on a new Windows platform. Everything seemed to installed OK. I then downloaded the latest Tinn-R (2.3.2.3 I think) and as I have always done I selected R - Configure - Permanent. I was greeted with a dialog box asking me for a mirror site. I don't remember this prompt before but I decided to play along and I select a mrror site. Then the process takes off installing what appears to be every package that has ever been conceived for R (translate - alot of packages). Is this normal? I repeated this about three times because at the end it gave me an error indicating sus-and-such library was not found. Each time it was a different library that couldn't be found. Finally on the third try it seemed to complete without error. Again I have never had to go through so many steps to get Tinn-R installed and configured. Did I do something wrong? Is there a bug in this package or a problem with the integration between R 2.9.2 and Tin-R 2.3.2.3? Thank you. Kevin __ 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] Tinn-R setup
I recently installed R 2.9.2 on a new Windows platform. Everything seemed to installed OK. I then downloaded the latest Tinn-R (2.3.2.3 I think) and as I have always done I selected R - Configure - Permanent. I was greeted with a dialog box asking me for a mirror site. I don't remember this prompt before but I decided to play along and I select a mrror site. Then the process takes off installing what appears to be every package that has ever been conceived for R (translate - alot of packages). Is this normal? I repeated this about three times because at the end it gave me an error indicating sus-and-such library was not found. Each time it was a different library that couldn't be found. Finally on the third try it seemed to complete without error. Again I have never had to go through so many steps to get Tinn-R installed and configured. Did I do something wrong? Is there a bug in this package or a problem with the integration between R 2.9.2 and Tin-R 2.3.2.3? Thank you. Kevin __ 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 residuals function.
I have a misunderstanding on the residuals function in 'R'. In the stats package the residuals for the output of a HoltWinters fit is residuals.HoltWinters and the source looks like: stats:::residuals.HoltWinters function (object, ...) object$x - object$fitted[, 1] environment: namespace:stats If I execute the following code (I only include the forecast package for the data set): library(forecast) library(expsmooth) library(fma) f - HoltWinters(wineind) This results in a HoltWinters fit to the data in the variable 'f'. If I look at the first few data points: f$x[1:10] [1] 15136 16733 20016 17708 18019 19227 22893 23739 21133 22591 And at the fitted values: f$fitted[1:10,1] [1] 14043.33 16849.90 18885.04 20474.27 18836.84 21665.11 24118.32 25198.27 [9] 22901.40 24450.89 Then at the residuals: residuals(f)[1:10] [1] 984.6741 1127.0982 1122.9566 879.7321 661.1564 459.8873 [7] 1698.6798 3580.7276 -1941.4028 -2196.8865 Take the first element. It is clearly not simply f$x[1] - f$fitted[1,1] (which is 1092.674 not 984.6741). So my question is, What is actually being subtracted to get the residuals? Thank you. Kevin __ 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] predict.HoltWinters source misunderstanding.
If I look in the stats package for the 'R' source code for predict.HoltWinters I see the following lines: vars - function(h) { psi - function(j) object$alpha * (1 + j * object$beta) + (j%%f == 0) * object$gamma * (1 - object$alpha) var(residuals(object)) * if (object$seasonal == additive) sum(1, (h 1) * sapply(1L:(h - 1), function(j) crossprod(psi(j There is more source code but my question is on the call to crossprod(psi(j)). Looking at the psi function I could see no elements that are vectors (or matrices) or that result in such. In fact if I replace the call to crossprod with: sum(1, (h 1) * sapply(1L:(h - 1), function(j) { r - psi(j) r*r } )) I get the same results. If I further augment my test by printing out the length of the return value from 'psi' I can see no case where 'psi' would return anything but a single number. So my question is, why the call to crossprod when simply squaring the number will do. Am I missing a case when the members of the HoltWinters fit would be vectors and then crossprod would be appropriate or is this just an oversight? Thank you. Kevin __ 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] Building 'R' from source for Windows.
Thank you for looking into this. It turns out the problem was You are misinterpreting R_HOME. . . I thought R_HOME was were I installed R not the directory where I was trying to compile the source. Once I moved the extra stuff that RTools.exe installed in what I thought was the R installation directory to where I was trying to cimpile the R source it became apparent what I had done wrong. Thanks again. But unfortunately this brings up two more questions. One the function or project that I would like to start debugging is in appl (I would like to step into the L-LBFGSB lbfgsb.c code). The documentation that was with the RTools code mentioned that if I built a package like 'make DEBUG=T package' it would insert -gdwarf-2 in the compiler swtiches and that would allow be to set a breakpoint. Well I am not sure if the fact that this is not a package if that is why it didn't work but if I go to the appl directory and enter 'make DEBUG=T', I don't see gcc called with -gdwarf-2. If I go to the gnuwin32 directory and try to build all of 'R' like 'make DEBUG=T all recommended' I see the -gdwarf=2 flag added to compilation of each file. But I would rather not add debugging information to all the source in R. Any suggestions? The second question is kind of like 'Where do I go from here?'. From the instructions I get that I probably need to use 'Inno' to build an installation. Hopefully once the first problem is solved this installation will have debugging symbols where I need them. If I run the resultant self-extracting installer will that just overwrite my R installation and now I debug with that? Again, thanks for the additional tips. It has been a long time since I last debugged with anything other than Windows Visual Studio and this will take some getting used to. Kevin Peter Dalgaard p.dalga...@biostat.ku.dk wrote: rkevinbur...@charter.net wrote: I know I am going to catch alot of comments for this question but I am really stuck. If there is some written documentation that I have missed please redirect me. I want to build 'R' from source on a Windows Platform. The main reasons are that I want to check out a debugging some existing packages so I need to build with debug symbols and I want to check out a 64-bit version of 'R'. So I read the instuructions and downloaded and installed 'rtools' and extracted the source. Then I ran into this statement in R-admin.pdf: Open a command window at ‘R_HOME/src/gnuwin32’. Edit ‘MkRules’ to set the appropriate paths as needed and to set the type(s) of help that you want built. Beware: ‘MkRules’ contains tabs and some editors (e.g., WinEdt) silently remove them. Then run make all recommended and sit back and wait while the basic compile takes place. But when I go to this directory I don't see MkRules. In fact I don't see any files, just folders (bitmap and unicode). Are the instructions wrong ? Have I missed a step? Or is there somewhere I can retrieve the missing file (MkRules)? Thank you. Kevin It should be there. Three possibilities: - Your editor is not showing it because of extension issues (look for All files) - You are misinterpreting R_HOME as something other than the source directory. (Could this be a typo? R_HOME is usually the destination dir, but source is what makes sense here. Or are the instructions assuming builddir=srcdir=destdir?) - You unpacked it incorrectly or got the wrong source file. I checked http://cran.r-project.org/src/base/R-2/R-2.9.1.tar.gz and it does have the file in the right place. -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - (p.dalga...@biostat.ku.dk) FAX: (+45) 35327907 __ 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] Building 'R' from source for Windows.
I know I am going to catch alot of comments for this question but I am really stuck. If there is some written documentation that I have missed please redirect me. I want to build 'R' from source on a Windows Platform. The main reasons are that I want to check out a debugging some existing packages so I need to build with debug symbols and I want to check out a 64-bit version of 'R'. So I read the instuructions and downloaded and installed 'rtools' and extracted the source. Then I ran into this statement in R-admin.pdf: Open a command window at ‘R_HOME/src/gnuwin32’. Edit ‘MkRules’ to set the appropriate paths as needed and to set the type(s) of help that you want built. Beware: ‘MkRules’ contains tabs and some editors (e.g., WinEdt) silently remove them. Then run make all recommended and sit back and wait while the basic compile takes place. But when I go to this directory I don't see MkRules. In fact I don't see any files, just folders (bitmap and unicode). Are the instructions wrong ? Have I missed a step? Or is there somewhere I can retrieve the missing file (MkRules)? Thank you. Kevin __ 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] Browser and Debug?
If I am on Windows and don't have GDB it does'nt look like it is possible. Any tips? Kevin roger koenker rkoen...@uiuc.edu wrote: At points of total desperation you can always consider the time-honored, avuncular advice -- RTFM, in this case Section 4.4 of Writing R Extensions. url:www.econ.uiuc.edu/~rogerRoger Koenker emailrkoen...@uiuc.eduDepartment of Economics vox: 217-333-4558University of Illinois fax: 217-244-6678Urbana, IL 61801 On Aug 13, 2009, at 10:20 AM, rkevinbur...@charter.net wrote: This may be asking more than can be reasonably expected. But, any tips on debugging the 'C' and Fortran code without trying to put together all the tools to compile from source? Kevin Erik Iverson eiver...@nmdp.org wrote: This article might help: http://www.biostat.jhsph.edu/~rpeng/docs/R-debug-tools.pdf -Original Message- From: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org ] On Behalf Of Inchallah Yarab Sent: Thursday, August 13, 2009 9:40 AM To: r-help@r-project.org Subject: [R] Browser and Debug? Hi, Someone can explain to me how use Browser and Debug ? thank you [[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-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-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] if confusion
Simple question: Why doesn't the following work? Or what 'R' rule am I missing? tclass - Testing 1 2 3 if(tclass == Testing 1 2 3) { cat(Testing, tclass, \n) } else { cat(tclass, \n) } I get an error 'else' is unexpected. Thank you. Kevin __ 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] Simple cat statement - output truncated
So then I am to assume that the output of 'cat' can be truncated by passing it bad arrays. That is the only difference between the reproducible code you show and mine. It is just a theory but say that the components array is not dimmensioned for 4 elements. It seems a little strange if that is the case that a reference error is not thrown and just the output of the cat call is affected. Kevin Duncan Murdoch murd...@stats.uwo.ca wrote: On 7/15/2009 9:53 AM, rkevinbur...@charter.net wrote: I have a statement: cat(myforecast ETS(, paste(object$components[1], object$components[2], object$components[3], object$components[4], sep = ,), ) , n, \n) That generates: cast ETS( A,N,N,FALSE ) 3 Anyone guess as to why the first 5 letters are truncated/missing? You are probably being punished for posting non-reproducible code*. When I try a reproducible version of the line above, things look fine: cat(myforecast ETS(, paste(A,N,N,FALSE, sep = ,), ) , 3, \n) myforecast ETS( A,N,N,FALSE ) 3 Duncan Murdoch * R has a new predictive punishment module. It punishes you for things it knows you will do later. __ 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] Simple cat statement - output truncated
It has to be related to 'cat' because the output of 'cat' is truncated. I am just tyring to find out some possible reasons as to why it is truncated. I have been unable to form an array like is in the test program. Do you think there is something else that is gobbling up the output from cat that would make it appear to be truncated? Kevin Duncan Murdoch murd...@stats.uwo.ca wrote: On 7/16/2009 10:21 AM, rkevinbur...@charter.net wrote: So then I am to assume that the output of 'cat' can be truncated by passing it bad arrays. I certainly wouldn't draw that conclusion. Without a reproducible example, my assumption would be that it is unrelated to cat(). Duncan Murdoch That is the only difference between the reproducible code you show and mine. It is just a theory but say that the components array is not dimmensioned for 4 elements. It seems a little strange if that is the case that a reference error is not thrown and just the output of the cat call is affected. Kevin Duncan Murdoch murd...@stats.uwo.ca wrote: On 7/15/2009 9:53 AM, rkevinbur...@charter.net wrote: I have a statement: cat(myforecast ETS(, paste(object$components[1], object$components[2], object$components[3], object$components[4], sep = ,), ) , n, \n) That generates: cast ETS( A,N,N,FALSE ) 3 Anyone guess as to why the first 5 letters are truncated/missing? You are probably being punished for posting non-reproducible code*. When I try a reproducible version of the line above, things look fine: cat(myforecast ETS(, paste(A,N,N,FALSE, sep = ,), ) , 3, \n) myforecast ETS( A,N,N,FALSE ) 3 Duncan Murdoch * R has a new predictive punishment module. It punishes you for things it knows you will do later. __ 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] Simple cat statement - output truncated
I have a statement: cat(myforecast ETS(, paste(object$components[1], object$components[2], object$components[3], object$components[4], sep = ,), ) , n, \n) That generates: cast ETS( A,N,N,FALSE ) 3 Anyone guess as to why the first 5 letters are truncated/missing? Kevin __ 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] Repeated SANN values.
I tried optim using the SANN algoithm. To start things out I tried the example of solving the traveling salesman problem as given in the documentation. The example works just fine. But if I comment out the line: set.seed(123) # chosen to get a good soln relatively quickly More often than not it doesn't converge to the optimum solution as shown in the example. Alos with trace on it seems that the algoritm is easily fooled by a local mimimum as once it gets close to the solution it seems to get stuck and repeatedly returns the same value: A sample run: sann objective function values initial value 29625.00 iter 5000 value 13972.00 iter1 value 13501.00 iter15000 value 13501.00 iter2 value 13501.00 iter25000 value 13487.00 iter2 value 13487.00 final value 13487.00 sann stopped after 2 iterations Not that familiiar with the algoritmn Is that just a drawback of the algorithm or can I adjust the anealling temperature (temp) or the maximum temperature (tmax) or even the mximum number of iterations to kick it out of what appears to be a local minimum? I am willing to sacrifice extra compute time for better accuracy. Kevin __ 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] Need help in calculating studentized residuals/leverage values of non-linear model [nls()]
Is the output of residuals() the studentized residuals or just the residuals? Dieter Menne dieter.me...@menne-biomed.de wrote: Giam Xingli giam at nus.edu.sg writes: I hope I can get advice regarding the calculation of leverage values or studentized residual values of a non-linear regression model. It seems like rstudent() does not work on a nls object. residuals() should work for nls. Dieter __ 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] Souce macros help
I was trying to understand some of the source in optimi.c and in the SANN source I see: SETCADR(OS-R_gcall, x); PROTECT_WITH_INDEX(s = eval(OS-R_gcall, OS-R_env), ipx); REPROTECT(s = coerceVector(s, REALSXP), ipx); if(LENGTH(s) != n) error(_(candidate point in optim evaluated to length %d not %d), LENGTH(s), n); I think I need a little help it it is not too much to ask. Admitedly I could search for the definition of each of these macros and after some time decipher their meaning. But in an effort to save some time I am appealing to this group. First, SETCADR(OS-R_gcall, x); I was unable to find thie function call in the sources it seems to be called everywhere so searches turn up many hits. Second, PROTECT_WITH_INDEX(s = eval(OS-R_gcall, OS-R_env), ipx); I am assuming that this acutall calls the function pointed to by R_gcall. But, I am not sure where ipx fits in and what PROTECT_WITH_INDEX does. I read the wirting exstensions documentation and this specific macro is dealt with in section 5.9.1 but I still am having a hard time understanding what is happening. Perhaps the added call to eval is confiusing me. I am assuing that this makes the function call. Third, REPROTECT(s = coerceVector(s, REALSXP), ipx); Again the vector ipx shows up. Would someone please help me to understand what this statement is doing? Thank you for your time and patience. Kevin __ 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] embed?
I have a question on the function 'embed'. I ran the example x - 1:10 embed(x, dimension=3) This gives the output: [,1] [,2] [,3] [1,]321 [2,]432 [3,]543 [4,]654 [5,]765 [6,]876 [7,]987 [8,] 1098 I don't quite understand the output and why it is useful. First, there are only 8 rows down from 10 and the first element starts with 3. Of course I can think of explanations as to what is occuring but I cannot see how this is useful. I am sure it has application as i see this command used in much of the source but I just cannot see it now. The documentation states: Each row of the resulting matrix consists of sequences x[t], x[t-1], ..., x[t-dimension+1], where t is the original index of x. If x is a matrix, i.e., x contains more than one variable, then x[t] consists of the tth observation on each variable. This explanation doesn't seem to account for the dimension argument. Thank you for your comments. Kevin __ 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] Constrined dependent optimization.
I have decided to use this SANN approach to my problem but to keep the run time reasonable instead of 20,000 variables I will randomly sample this space to get the number of variables under 100. But I want to do this a number of times. Is there someone who could help me set up WINBUGS to repeat this optimization N times each time randomly picking 100 of the possible 20,000. Comments? Kevin Paul Smith phh...@gmail.com wrote: Apparently, the convergence is faster if one uses this new swap function: swapfun - function(x,N=100) { loc - c(sample(1:(N/2),size=1,replace=FALSE),sample((N/2):100,1)) tmp - x[loc[1]] x[loc[1]] - x[loc[2]] x[loc[2]] - tmp x } It seems that within 20 millions of iterations, one gets the exact optimal solution, which does not take too long. Paul On Mon, Mar 30, 2009 at 5:11 PM, Paul Smith phh...@gmail.com wrote: Optim with SANN also solves your example: --- f - function(x) sum(c(1:50,50:1)*x) swapfun - function(x,N=100) { loc - sample(N,size=2,replace=FALSE) tmp - x[loc[1]] x[loc[1]] - x[loc[2]] x[loc[2]] - tmp x } N - 100 opt1 - optim(fn=f,par=sample(1:N,N),gr=swapfun,method=SANN,control=list(maxit=5,fnscale=-1,trace=10)) opt1$par opt1$value --- We need to specify a large number of iterations to get the optimal solution. The objective function at the optimum is 170425, and one gets a close value with optim and SANN. Paul On Mon, Mar 30, 2009 at 2:22 PM, Hans W. Borchers hwborch...@googlemail.com wrote: Image you want to minimize the following linear function f - function(x) sum( c(1:50, 50:1) * x / (50*51) ) on the set of all permutations of the numbers 1,..., 100. I wonder how will you do that with lpSolve? I would simply order the coefficients and then sort the numbers 1,...,100 accordingly. I am also wondering how optim with SANN could be applied here. As this is a problem in the area of discrete optimization resp. constraint programming, I propose to use an appropriate program here such as the free software Bprolog. I would be interested to learn what others propose. Of course, if we don't know anything about the function f then it amounts to an exhaustive search on the 100! permutations -- probably not a feasible job. Regards, Hans Werner Paul Smith wrote: On Sun, Mar 29, 2009 at 9:45 PM, rkevinbur...@charter.net wrote: I have an optimization question that I was hoping to get some suggestions on how best to go about sovling it. I would think there is probably a package that addresses this problem. This is an ordering optimzation problem. Best to describe it with a simple example. Say I have 100 bins each with a ball in it numbered from 1 to 100. Each bin can only hold one ball. This optimization is that I have a function 'f' that this array of bins and returns a number. The number returned from f(1,2,3,4) would return a different number from that of f(2,1,3,4). The optimization is finding the optimum order of these balls so as to produce a minimum value from 'f'.I cannot use the regular 'optim' algorithms because a) the values are discrete, and b) the values are dependent ie. when the variable representing the bin location is changed (in this example a new ball is put there) the existing ball will need to be moved to another bin (probably swapping positions), and c) each variable is constrained, in the example above the only allowable values are integers from 1-100. So the problem becomes finding the optimum order of the balls. Any suggestions? If your function f is linear, then you can use lpSolve. Paul __ 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. -- View this message in context: http://www.nabble.com/Constrined-dependent-optimization.-tp22772520p22782922.html Sent from the R help mailing list archive at Nabble.com. __ 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-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting
[R] data.frame to array?
I have a list of data.frames str(bins) List of 19217 $ 100026:'data.frame': 1 obs. of 6 variables: ..$ Sku : chr 100026 ..$ Bin : chr T149C ..$ Count: int 108 ..$ X: int 20 ..$ Y: int 149 ..$ Z: chr 3 $ 100030:'data.frame': 1 obs. of 6 variables: ... As you can see one 'column' is Count. This list seems to contain 19217 data.frames. I would like to create an array of 19217 integers which hold the values of the Count column. I have tried the obvious (to me): bins[[1:3]]$Count But that returns NULL instead of an array of length 3 that I was expecting. Interestingly bins[[1]]$Count returns the first Count in the list of data frames. How do I get all of the Counts? Thank you. Kevin __ 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] Constrined dependent optimization.
Sorry I sent a description of the function I was trying to minimize but I must not have sent it to this group (and you). Hopefully with this clearer description of my problem you might have some suggestions. It is basically a warehouse placement problem. You have a warehouse that has many items each placed in a certain bin (the real warehouse has about 20,000 of these bins, hence the large number of variables that I want to input to optimize). Now assume that an order comes in for three items A, B, and C. In the worst case A will be on one end of the warehouse, B in the middle and C on the other end of the warehouse. The work involved in getting these items to fulfill this order is roughly proportional to the distance from A to B plus the distance from B to C (assuming the absolute positions are sorted). So the cost for fulfilling this order is this distance. In the ideal world A, B, and C would be right next to each other and the cost/distance would be minimized. So the function I want to minimize would be placing these 20,000 items in such a way so that the minimum work is involved in fulfilling the orders for the past month or two. Clearer? I can see that I may need to cut back on the variables 20,000 is probably too many. Maybe I can take the top 1,000 or so. I just am not sure of the packages available what to reasonably expect. I would like this optimization to complete in a reasonable amount of time (less than a few days). I have heard that SANN is slower than other optimization methods but it does have the feature of supplying a gradient as you pointed out. Are there other packages out there that might be better suited to such a large scale optimizaiton? Thanks again. Kevin Paul Smith phh...@gmail.com wrote: As I told you before, without knowing the definition of your function f, one cannot help much. Paul On Wed, Apr 1, 2009 at 3:15 PM, rkevinbur...@charter.net wrote: Thank you I had not considered using gradient in this fashion. Now as an add on question. You (an others) have suggested using SANN. Does your answer change if instead of 100 variables or bins there are 20,000? From the documentation L-BFGS-B is designed for a large number of variables. But maybe SANN can handle this as well. Kevin Paul Smith phh...@gmail.com wrote: Apparently, the convergence is faster if one uses this new swap function: swapfun - function(x,N=100) { loc - c(sample(1:(N/2),size=1,replace=FALSE),sample((N/2):100,1)) tmp - x[loc[1]] x[loc[1]] - x[loc[2]] x[loc[2]] - tmp x } It seems that within 20 millions of iterations, one gets the exact optimal solution, which does not take too long. Paul On Mon, Mar 30, 2009 at 5:11 PM, Paul Smith phh...@gmail.com wrote: Optim with SANN also solves your example: --- f - function(x) sum(c(1:50,50:1)*x) swapfun - function(x,N=100) { loc - sample(N,size=2,replace=FALSE) tmp - x[loc[1]] x[loc[1]] - x[loc[2]] x[loc[2]] - tmp x } N - 100 opt1 - optim(fn=f,par=sample(1:N,N),gr=swapfun,method=SANN,control=list(maxit=5,fnscale=-1,trace=10)) opt1$par opt1$value --- We need to specify a large number of iterations to get the optimal solution. The objective function at the optimum is 170425, and one gets a close value with optim and SANN. Paul On Mon, Mar 30, 2009 at 2:22 PM, Hans W. Borchers hwborch...@googlemail.com wrote: Image you want to minimize the following linear function f - function(x) sum( c(1:50, 50:1) * x / (50*51) ) on the set of all permutations of the numbers 1,..., 100. I wonder how will you do that with lpSolve? I would simply order the coefficients and then sort the numbers 1,...,100 accordingly. I am also wondering how optim with SANN could be applied here. As this is a problem in the area of discrete optimization resp. constraint programming, I propose to use an appropriate program here such as the free software Bprolog. I would be interested to learn what others propose. Of course, if we don't know anything about the function f then it amounts to an exhaustive search on the 100! permutations -- probably not a feasible job. Regards, Hans Werner Paul Smith wrote: On Sun, Mar 29, 2009 at 9:45 PM, rkevinbur...@charter.net wrote: I have an optimization question that I was hoping to get some suggestions on how best to go about sovling it. I would think there is probably a package that addresses this problem. This is an ordering optimzation problem. Best to describe it with a simple example. Say I have 100 bins each with a ball in it numbered from 1 to 100. Each bin can only hold one ball. This optimization is that I have a function 'f' that this array of bins and
Re: [R] Constrined dependent optimization.
Thank you I had not considered using gradient in this fashion. Now as an add on question. You (an others) have suggested using SANN. Does your answer change if instead of 100 variables or bins there are 20,000? From the documentation L-BFGS-B is designed for a large number of variables. But maybe SANN can handle this as well. Kevin Paul Smith phh...@gmail.com wrote: Apparently, the convergence is faster if one uses this new swap function: swapfun - function(x,N=100) { loc - c(sample(1:(N/2),size=1,replace=FALSE),sample((N/2):100,1)) tmp - x[loc[1]] x[loc[1]] - x[loc[2]] x[loc[2]] - tmp x } It seems that within 20 millions of iterations, one gets the exact optimal solution, which does not take too long. Paul On Mon, Mar 30, 2009 at 5:11 PM, Paul Smith phh...@gmail.com wrote: Optim with SANN also solves your example: --- f - function(x) sum(c(1:50,50:1)*x) swapfun - function(x,N=100) { loc - sample(N,size=2,replace=FALSE) tmp - x[loc[1]] x[loc[1]] - x[loc[2]] x[loc[2]] - tmp x } N - 100 opt1 - optim(fn=f,par=sample(1:N,N),gr=swapfun,method=SANN,control=list(maxit=5,fnscale=-1,trace=10)) opt1$par opt1$value --- We need to specify a large number of iterations to get the optimal solution. The objective function at the optimum is 170425, and one gets a close value with optim and SANN. Paul On Mon, Mar 30, 2009 at 2:22 PM, Hans W. Borchers hwborch...@googlemail.com wrote: Image you want to minimize the following linear function f - function(x) sum( c(1:50, 50:1) * x / (50*51) ) on the set of all permutations of the numbers 1,..., 100. I wonder how will you do that with lpSolve? I would simply order the coefficients and then sort the numbers 1,...,100 accordingly. I am also wondering how optim with SANN could be applied here. As this is a problem in the area of discrete optimization resp. constraint programming, I propose to use an appropriate program here such as the free software Bprolog. I would be interested to learn what others propose. Of course, if we don't know anything about the function f then it amounts to an exhaustive search on the 100! permutations -- probably not a feasible job. Regards, Hans Werner Paul Smith wrote: On Sun, Mar 29, 2009 at 9:45 PM, rkevinbur...@charter.net wrote: I have an optimization question that I was hoping to get some suggestions on how best to go about sovling it. I would think there is probably a package that addresses this problem. This is an ordering optimzation problem. Best to describe it with a simple example. Say I have 100 bins each with a ball in it numbered from 1 to 100. Each bin can only hold one ball. This optimization is that I have a function 'f' that this array of bins and returns a number. The number returned from f(1,2,3,4) would return a different number from that of f(2,1,3,4). The optimization is finding the optimum order of these balls so as to produce a minimum value from 'f'.I cannot use the regular 'optim' algorithms because a) the values are discrete, and b) the values are dependent ie. when the variable representing the bin location is changed (in this example a new ball is put there) the existing ball will need to be moved to another bin (probably swapping positions), and c) each variable is constrained, in the example above the only allowable values are integers from 1-100. So the problem becomes finding the optimum order of the balls. Any suggestions? If your function f is linear, then you can use lpSolve. Paul __ 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. -- View this message in context: http://www.nabble.com/Constrined-dependent-optimization.-tp22772520p22782922.html Sent from the R help mailing list archive at Nabble.com. __ 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-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
Re: [R] Constrined dependent optimization.
It would in the stictess sense be non-linear since it is only defined for descrete interface values for each variable. And in general it would be non-linear anyway. If I only have three variables which can take on values 1,2,3 then f(1,2,3) could equal 0 and f(2,1,3) could equal 10. Thank you for the suggestions. Kevin Paul Smith phh...@gmail.com wrote: On Sun, Mar 29, 2009 at 9:45 PM, rkevinbur...@charter.net wrote: I have an optimization question that I was hoping to get some suggestions on how best to go about sovling it. I would think there is probably a package that addresses this problem. This is an ordering optimzation problem. Best to describe it with a simple example. Say I have 100 bins each with a ball in it numbered from 1 to 100. Each bin can only hold one ball. This optimization is that I have a function 'f' that this array of bins and returns a number. The number returned from f(1,2,3,4) would return a different number from that of f(2,1,3,4). The optimization is finding the optimum order of these balls so as to produce a minimum value from 'f'.I cannot use the regular 'optim' algorithms because a) the values are discrete, and b) the values are dependent ie. when the variable representing the bin location is changed (in this example a new ball is put there) the existing ball will need to be moved to another bin (probably swapping positions), and c) each variable is constrained, in the example above the only allowable values are integers from 1-100. So the problem becomes finding the optimum order of the balls. Any suggestions? If your function f is linear, then you can use lpSolve. Paul __ 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] Constrined dependent optimization.
I have an optimization question that I was hoping to get some suggestions on how best to go about sovling it. I would think there is probably a package that addresses this problem. This is an ordering optimzation problem. Best to describe it with a simple example. Say I have 100 bins each with a ball in it numbered from 1 to 100. Each bin can only hold one ball. This optimization is that I have a function 'f' that this array of bins and returns a number. The number returned from f(1,2,3,4) would return a different number from that of f(2,1,3,4). The optimization is finding the optimum order of these balls so as to produce a minimum value from 'f'.I cannot use the regular 'optim' algorithms because a) the values are discrete, and b) the values are dependent ie. when the variable representing the bin location is changed (in this example a new ball is put there) the existing ball will need to be moved to another bin (probably swapping positions), and c) each variable is constrained, in the example above the only allowable values are integers from 1-100. So the problem becomes finding the optimum order of the balls. Any suggestions? Kevin __ 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] Bug? FORTTRAN help
I was feeling masochistic the other day and we have been having some wierd memory problems so I started digging into the source for L-BFGS-B. In the lbgfsb.c file I see the following code: /* Cholesky factorization of (2,2) block of wn. */ F77_CALL(dpofa)(wn[*col + 1 + (*col + 1) * wn_dim1], m2, col, info); if (*info != 0) { *info = -2; return; } If I am not mistaken this says that there is a m2 * col matrix that starts at 'col + 1 + (col + 1) * wn_dm1. Where wn_dm1 is 2 * m. My first question is to verify that statement. Say I am trying to optimize the banana function as given in the documentation. In that case n = 2 and the default m = 5. So m2 is 10 and wn_dim1 is 20 and the dimension of wn is 100 (this is all by deduction. So if col is 5 then the offset into the array is 55 and there is not room in the vector for a 10 x 5 array. I am worried that the optimizer will silently write info memory that it shouldn't but more than likely it is something that I don't understand. So please vefify my first statement. Thank you. Kevin __ 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] .Internal
Sorry to be so dense but the article that you suggest does not give any information on how the arguments are packed up. I look at the call: val - .Internal(fmin(function(arg) -f(arg, ...), lower, upper, tol)) and then with the help of this article I find do_fmin in optimize.c: SEXP attribute_hidden do_fmin(SEXP call, SEXP op, SEXP args, SEXP rho) Again there doesn't seem to be any coorespondance between lower, upper, tol and the arguments to do_fmin. So I am missing a step. Thank you. Kevin Berwin A Turlach ber...@maths.uwa.edu.au wrote: G'day Kevin, On Wed, 18 Mar 2009 21:46:51 -0700 rkevinbur...@charter.net wrote: I was trying to find source for optimize and I ran across function (f, interval, ..., lower = min(interval), upper = max(interval), maximum = FALSE, tol = .Machine$double.eps^0.25) { if (maximum) { val - .Internal(fmin(function(arg) -f(arg, ...), lower, upper, tol)) list(maximum = val, objective = f(val, ...)) } else { val - .Internal(fmin(function(arg) f(arg, ...), lower, upper, tol)) list(minimum = val, objective = f(val, ...)) } } Then I did a search for fmin and i came up with: /* fmin(f, xmin, xmax tol) */ SEXP attribute_hidden do_fmin(SEXP call, SEXP op, SEXP args, SEXP rho) So my question is where do I find the intermediary step between .Internal(fmin(function(arg) f(arg, ...), lower, upper, tol)) and SEXP attribute_hidden do_fmin(SEXP call, SEXP op, SEXP args, SEXP rho) @Article{Rnews:Ligges:2006, author = {Uwe Ligges}, title= {{R} {H}elp {D}esk: {Accessing} the Sources}, journal = {R News}, year = 2006, volume = 6, number = 4, pages= {43--45}, month= {October}, url = http, pdf= Rnews2006-4 } http://CRAN.R-project.org/doc/Rnews/Rnews_2006-4.pdf The number of arguments doesn't match up. I am guessing that lower and upper somehow get merged into the args. And rho is 'tol'. Right? Unlikely. In Writing R Extensions (and the functions I looked up), 'rho' usually denotes an environment that is used to evaluate expressions in. Typically (i.e. in cases that I had need to look at), all arguments are rolled into the SEXP arg for internal functions. HTH. Cheers, Berwin __ 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] variance/mean
At the risk of appearing ignorant why is the folowing true? o - cbind(rep(1,3),rep(2,3),rep(3,3)) var(o) [,1] [,2] [,3] [1,]000 [2,]000 [3,]000 and mean(o) [1] 2 How do I get mean to return an array similar to var? I would expect in the above example a vector of length 3 {1,2,3}. Thank you for your help. Kevin __ 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] .Internal
I was trying to find source for optimize and I ran across function (f, interval, ..., lower = min(interval), upper = max(interval), maximum = FALSE, tol = .Machine$double.eps^0.25) { if (maximum) { val - .Internal(fmin(function(arg) -f(arg, ...), lower, upper, tol)) list(maximum = val, objective = f(val, ...)) } else { val - .Internal(fmin(function(arg) f(arg, ...), lower, upper, tol)) list(minimum = val, objective = f(val, ...)) } } Then I did a search for fmin and i came up with: /* fmin(f, xmin, xmax tol) */ SEXP attribute_hidden do_fmin(SEXP call, SEXP op, SEXP args, SEXP rho) So my question is where do I find the intermediary step between .Internal(fmin(function(arg) f(arg, ...), lower, upper, tol)) and SEXP attribute_hidden do_fmin(SEXP call, SEXP op, SEXP args, SEXP rho) The number of arguments doesn't match up. I am guessing that lower and upper somehow get merged into the args. And rho is 'tol'. Right? Thank you for your comments. Kevin __ 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] sweep?
I am having a hard time understanding just what 'sweep' does. The documentation states: Return an array obtained from an input array by sweeping out a summary statistic. So what does it mean weeping out a summary statistic? Thank you. Kevin __ 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] Yet another large file question.
I am using the XML package to form a rather large XML file. It seems to go OK untill the file length gets larger than about 30Mb then it seems that the last tokens of the xml file are unmatched. It is almost like the file hasn't been flushed because the file is OK with the exception of the last lines. I was wondering two things 1) Has anyone else run into similar limitations with saveXML in the XML package and if there is a work-around? 2) Am I better off using more pimitive write methods to get the data to a file? If so any suggestions? Thank you. Kevin __ 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] updating packages?
I am trying to update the packages that I have installed but I get the following warning messages: package 'tseries' successfully unpacked and MD5 sums checked Warning: cannot remove prior installation of package 'tseries' bundle 'forecasting' successfully unpacked and MD5 sums checked Warning: cannot remove prior installation of package 'forecast' What does that mean? How can I update these packages? Thank you. Kevin __ 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] Reshape question.
This hopefully is trivial. I am trying to reshape the data using the reshape package. First I read in the data: a2009 - read.csv(Total2009.dat, header = TRUE) Then I trim it so that it only contains the columns that I have interested in: m2009 - melt(a2009, id.var=c(DayOfYear,Category,SubCategory,Sku), measure.var=c(Quantity), na.rm=TRUE) Then I start to formulate the data that I will process: c2009 - cast(m2009, DayOfYear ~ variable | Category, sum) Finally I aggregate the data: t2009 - cast(m2009, DayOfYear ~ variable, sum) My question is on the third step above (repeated here) c2009 - cast(m2009, DayOfYear ~ variable | Category, sum) This gets the data assocated with a unique 'Category' name. I want to get the data grouped by 'Category' and 'SubCategory'. The 'SubCategory' is not unique but the combination 'Category' and 'SubCategory' form a unique pair. What would be the formula that would give me the data grouped by Category AND SubCategory? Would it be as simple as: c2009 - cast(m2009, DayOfYear ~ variable | Category SubCategory, sum) ? Thank you for your suggestions. Kevin __ 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] Reshape question.
Thank you for you reply. I will try this. The inital few rows in the .dat file look like: Year,DayOfYear,Sku,Quantity,CatId,Category,SubCategory 2009,1,100051,1,10113,MEN,Historical men's 2009,1,100130,1,10638,ACCESSORIES MAKEUP,ALL Kids Accessories 2009,1,100916,1,10222,WOMEN,TV Movies Women 2009,1,101241,1,10897,HOLIDAY,Colonial (Presidents) 2009,1,101252,1,10640,ACCESSORIES MAKEUP,Finishing Touches 2009,1,101298,1,10865,HOLIDAY,Easter 2009,1,101613,1,10410,GIRLS,Classic Girls 2009,1,101645,1,10320,BOYS,Superheroes Boys 2009,1,101648,1,10320,BOYS,Superheroes Boys 2009,1,101718,1,10897,HOLIDAY,Colonial (Presidents) 2009,1,101719,1,10897,HOLIDAY,Colonial (Presidents) 2009,1,101751,1,10420,GIRLS,Superheroes Girls 2009,1,102125,1,10638,ACCESSORIES MAKEUP,ALL Kids Accessories 2009,1,102174,1,10897,HOLIDAY,Colonial (Presidents) 2009,1,102558,1,10636,ACCESSORIES MAKEUP,Armor/Weapons/Guns 2009,1,102582,1,10636,ACCESSORIES MAKEUP,Armor/Weapons/Guns 2009,1,102717,1,10862,HOLIDAY,Christmas 2009,1,104705,1,10518,PLUS,Plus Women 2009,1,104745,6,10748,HATS, WIGS MASKS,Wigs - Men's 2009,1,104745,1,10748,HATS, WIGS MASKS,Wigs - Men's 2009,1,104751,1,10310,BOYS,Classic Boys 2009,1,105238,1,10742,HATS, WIGS MASKS,Hats-Miscellaneous 2009,1,105352,10,10742,HATS, WIGS MASKS,Hats-Miscellaneous 2009,1,107420,10,10744,HATS, WIGS MASKS,Masks - Miscellaneous 2009,1,107420,1,10744,HATS, WIGS MASKS,Masks - Miscellaneous 2009,1,107479,1,10743,HATS, WIGS MASKS,Masks - Famous 2009,1,107479,1,10743,HATS, WIGS MASKS,Masks - Famous If your propose solution works I am confused as to why during the original I was able to specify: c2009 - cast(m2009, DayOfYear ~ variable | Category, sum) t2009 - cast(m2009, DayOfYear ~ variable, sum) By combining this into one 'cast' is it better (as in faster)? Thanks again. Kevin Tal Galili tal.gal...@gmail.com wrote: how about: c2009 - cast(m2009, Category + SubCategory +DayOfYear ~ variable , sum) ? p.s: toy data would be nice to have :) On Wed, Mar 11, 2009 at 9:47 PM, rkevinbur...@charter.net wrote: This hopefully is trivial. I am trying to reshape the data using the reshape package. First I read in the data: a2009 - read.csv(Total2009.dat, header = TRUE) Then I trim it so that it only contains the columns that I have interested in: m2009 - melt(a2009, id.var=c(DayOfYear,Category,SubCategory,Sku), measure.var=c(Quantity), na.rm=TRUE) Then I start to formulate the data that I will process: c2009 - cast(m2009, DayOfYear ~ variable | Category, sum) Finally I aggregate the data: t2009 - cast(m2009, DayOfYear ~ variable, sum) My question is on the third step above (repeated here) c2009 - cast(m2009, DayOfYear ~ variable | Category, sum) This gets the data assocated with a unique 'Category' name. I want to get the data grouped by 'Category' and 'SubCategory'. The 'SubCategory' is not unique but the combination 'Category' and 'SubCategory' form a unique pair. What would be the formula that would give me the data grouped by Category AND SubCategory? Would it be as simple as: c2009 - cast(m2009, DayOfYear ~ variable | Category SubCategory, sum) ? Thank you for your suggestions. Kevin __ 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. -- -- My contact information: Tal Galili Phone number: 972-50-3373767 FaceBook: Tal Galili My Blogs: www.talgalili.com www.biostatistics.co.il __ 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] hatvalues?
I am struiggling a bit with this function 'hatvalues'. I would like a little more undrestanding than taking the black-box and using the values. I looked at the Fortran source and it is quite opaque to me. So I am asking for some help in understanding the theory. First, I take the simplest case of a single variant. For this I turn o John Fox's book, Applied Regression Analysis and Generalized Linear Models, p 245 and generate this 'R' code: library(car) attach(Davis) # remove the NA's narepwt - repwt[!is.na(repwt)] meanrw - mean(narepwt) drw - narepwt - meanrw ssrw - sum(drw * drw) h - 1/length(narepwt) + (drw * drw)/ssrw h This gives me a array of values the largest of which is order(h, decreasing=TRUE) [1] 21 52 17 93 30 62 158 113 175 131 182 29 106 125 123 146 91 99 So the largest hatvalue is h[21] [1] 0.1041207 Which doesn't match the 0.714 value that is reported in the book but I will probably take that up with the author later. Then I use more of 'R' and I get fit - lm(weight ~ repwt) hr - hatvalues(fit) hr[21] 21 0.1041207 So this matches which is reasusing. My question is this, given the QR transformation and the residuals derived from that transformation what is a simple matrix formula for the hatvalues? From http://en.wikipedia.org/wiki/Linear_regression I get residuals = y - Hy = y(I - H) or H = -(residuals/y - I) fit - lm(weight ~ repwt) h - -(residuals(fit)/weight[as.numeric(names(residuals(fit)))] - diag(1,length(residuals(fit)), length(residuals(fit This generates a matrix but I cannot see any coerrelation between this hat-matrix and the return from hatvalues. Comments? Thank you. Kevin __ 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] RES: object .trPaths not found
Yes. But I found out I the files permissions were set so that the configuration had a null effect and the configuration silently ignored the fact that the file could not be written to. Thank you. Kevin Leandro Marino lean...@cesgranrio.org.br wrote: Did you do the configuration of Tinn-R after the installation? Atenciosamente, Leandro Lins Marino Centro de Avaliação Fundação CESGRANRIO Rua Santa Alexandrina, 1011 - 2º andar Rio de Janeiro, RJ - CEP: 20261-903 R (21) 2103-9600 R.:236 ( (21) 8777-7907 ( lean...@cesgranrio.org.br Aquele que suporta o peso da sociedade é precisamente aquele que obtém as menores vantagens. (SMITH, Adam) Antes de imprimir pense em sua responsabilidade e compromisso com o MEIO AMBIENTE -Mensagem original- De: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org] Em nome de Uwe Ligges Enviada em: sábado, 28 de fevereiro de 2009 16:34 Para: rkevinbur...@charter.net Cc: R help Assunto: Re: [R] object .trPaths not found rkevinbur...@charter.net wrote: I am running an R script with Tinn-R (2.2.0.1) and I get the error message Error in source(.trPaths[4], echo = TRUE, max.deparse.length = 150) : object .trPaths not found Any solutions? Maybe, but the may depend on your script, your OS, your R version, used packages and so on. Hence please read and follow the posting guide and provide commented, minimal, self-contained, reproducible code. Uwe Ligges Thank you. Kevin __ 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-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] RES: object .trPaths not found
Yes. But I found out I the files permissions were set so that the configuration had a null effect and the configuration silently ignored the fact that the file could not be written to. Thank you. Kevin Leandro Marino lean...@cesgranrio.org.br wrote: Did you do the configuration of Tinn-R after the installation? Atenciosamente, Leandro Lins Marino Centro de Avaliação Fundação CESGRANRIO Rua Santa Alexandrina, 1011 - 2º andar Rio de Janeiro, RJ - CEP: 20261-903 R (21) 2103-9600 R.:236 ( (21) 8777-7907 ( lean...@cesgranrio.org.br Aquele que suporta o peso da sociedade é precisamente aquele que obtém as menores vantagens. (SMITH, Adam) Antes de imprimir pense em sua responsabilidade e compromisso com o MEIO AMBIENTE -Mensagem original- De: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org] Em nome de Uwe Ligges Enviada em: sábado, 28 de fevereiro de 2009 16:34 Para: rkevinbur...@charter.net Cc: R help Assunto: Re: [R] object .trPaths not found rkevinbur...@charter.net wrote: I am running an R script with Tinn-R (2.2.0.1) and I get the error message Error in source(.trPaths[4], echo = TRUE, max.deparse.length = 150) : object .trPaths not found Any solutions? Maybe, but the may depend on your script, your OS, your R version, used packages and so on. Hence please read and follow the posting guide and provide commented, minimal, self-contained, reproducible code. Uwe Ligges Thank you. Kevin __ 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-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] Gradient function for optim.
Thank you. I saw the source. But I am not sure how to get from .Internal(optim(...)) to fmingr. Kevin Katharine Mullen k...@few.vu.nl wrote: see the fmingr function in src/main/optim.c (https://svn.r-project.org/R/trunk/src/main/optim.c) On Wed, 25 Feb 2009 rkevinbur...@charter.net wrote: I have read that when the gradient function is not supplied (is null) then first order differencing is used to find the differential. I was trying to track down this for my own information but I run into .Internal(optim.). I was not sure where to look next to see the function that is automatically supplied for the gradient. Thank you. Kevin __ 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] Optim parscale?
I am not clear on what is happening with parscale in optim It seems that scaling the parameters will produce unpredictable results in a non-linear function (which is the purpose of optim right?) The documentation states: parscale A vector of scaling values for the parameters. Optimization is performed on par/parscale and these should be comparable in the sense that a unit change in any element produces about a unit change in the scaled value. I am not sure how it can be guaranteed that this will be case. I first saw a non-unity parscale in the ARIMA code where I see code like: ses - summary(fit)$coefficients[, 2] parscale - c(parscale, 10 * ses) Comments? Thank you. Kevin __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] object .trPaths not found
FYI. I found the solution. My RProfile.site file could not be written to because of permissions. Whan I selected the Configure/Permanent option in Tinn-R it was silently ignoring the fact that the file could not be written to. When I adjusted the permissions, all was well. Thank you. Kevin Kingsford Jones kingsfordjo...@gmail.com wrote: RSiteSearch('trpaths') will lead you to potential solutions... On Fri, Feb 27, 2009 at 7:18 PM, rkevinbur...@charter.net wrote: I am running an R script with Tinn-R (2.2.0.1) and I get the error message Error in source(.trPaths[4], echo = TRUE, max.deparse.length = 150) : object .trPaths not found Any solutions? Thank you. Kevin __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Variable scope.
I have a question on scope/reference/value type of variables with 'R'. The issue cam up first when I look at the arima code. I see code like: myupARIMA - function(mod, phi, theta) { . . . . mod } Then armafn - function(p, trans) { . . . . Z - upARIMA(mod, trarma[[1]], trarma[[2]]) . . . . res - .Call(R_ARIMA_Like, x, Z$phi, Z$theta, Z$Delta, Z$a, Z$P, Z$Pn, as.integer(0), FALSE) . . . . } The question is that ARIMA_Like will make changes to the arrays Z$P etc. Since upARIMA essentially returns 'mod' are changes to the arrays passed as Z$... to ARUINA_Like refkected in the original 'mod'? Or another way of phrasing the question is 'Z' a reference variable? Are the members such as Z$a also passed as reference hence changes to Z$a should also be seen in mod$a? In the arima code the references to x and mod seem to be essentially global variables as armafn is 'nested' in arima. Will changes to x also be reflected in the original x? (Like x - x - (xreg %*% par[narma + (1:ncxreg))) Thank you. Kevin __ 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] object .trPaths not found
I am running an R script with Tinn-R (2.2.0.1) and I get the error message Error in source(.trPaths[4], echo = TRUE, max.deparse.length = 150) : object .trPaths not found Any solutions? Thank you. Kevin __ 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] Gradient function for optim.
I have read that when the gradient function is not supplied (is null) then first order differencing is used to find the differential. I was trying to track down this for my own information but I run into .Internal(optim.). I was not sure where to look next to see the function that is automatically supplied for the gradient. Thank you. Kevin __ 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] Arima bug?
I was looking at the 'R' code associated with arima. I see the following: upARIMA - function(mod, phi, theta) { p - length(phi) q - length(theta) mod$phi - phi mod$theta - theta r - max(p, q + 1) if (p 0) mod$T[1:p, 1] - phi if (r 1) mod$Pn[1:r, 1:r] - .Call(R_getQ0, phi, theta) else if (p 0) mod$Pn[1, 1] - 1/(1 - phi^2) else mod$Pn[1, 1] - 1 mod$a[] - 0 mod } In particular the statement: else if (p 0) mod$Pn[1, 1] - 1/(1 - phi^2) If p (the length of the phi vector) is greater than one then it looks like the code is trying to assign a vector to a single element which will generate an error in 'R'. Much like x - 1:3 x[1] - 1:3 So I was wondering if there is something that I am missing that will guarantee that the length of the phi vector in this case is always 1. If this is not the case then it seems that this is a bug in the arima code. Comments? Kevin __ 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] Sign differences amoung QR solutions.
I was noticing mainly sign differences amoung the solutions to QR decomposition. For example R: x - matrix(c(12,-51,4,6,167,-68,-4,24,-41),nrow=3,byrow=T) x [,1] [,2] [,3] [1,] 12 -514 [2,]6 167 -68 [3,] -4 24 -41 r - qr(x) r$qr [,1] [,2] [,3] [1,] -14.000 -21.000 14 [2,] 0.4285714 -175.000 70 [3,] -0.28571430.1107692 -35 qr.Q(r) [,1] [,2][,3] [1,] -0.8571429 0.3942857 0.33142857 [2,] -0.4285714 -0.9028571 -0.03428571 [3,] 0.2857143 -0.1714286 0.94285714 When I feed this matrix into http://www.bluebit.gr/matrix-calculator/calculate.aspx I get: R -14.000 -21.000 14.000 0.000 -175.000 70.000 0.0000.000 -35.000 Q -0.857 0.394 0.331 -0.429 -0.903 -0.034 0.286 -0.171 0.943 Which seems to agree with the results of 'R' This same matrix on http://en.wikipedia.org/wiki/QR_decomposition R 14 -21 14 0 175 -70 0 0 35 Q 6/7 -69/175 -58/175 3/7 158/175 6/175 -2/7 6/36 -33/35 I have yet another algoithm that puts the R matrix as: -14 -21 14 0 -175 70 0 0 35 And Q as: Q: -0.8571 0.3943 -0.3314 -0.4286 -0.9029 0.0343 0.2857 -0.1714 -0.9429 So Wikipedia shows different signs but the property X = QR holds in all cases. My question is are there any gotchas (similar to the sign for tan) that I need to be aware of when using the results of a QR decomposition? With the sign differences it seems that there could be many possible decompositions for one matrix depending on the algoithem. Thank you. Kevin __ 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] Hessian?
I am new to 'R' and also new to the concept of a 'Hessian' with non-linear optimization. I would like to avoid going through all of the reference articles given with ?optim as access to a library is not handy. Would someone be able to elighten me on what is in the Hessian matrix if 'hessian = TRUE' with optim? The documentation indicates that this is a numerically differentiated Hessian. So I am assuming that some small delta is chosen. How is the delta chosen ? Thank you. Kevin __ 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] Foreign function call
Let me get more specific. I think it this can be answered then I can translate the information to other calls. In the arima 'R' code there is a reference to .Call(R_TSconv, a, b) If from the console I type: .Call(R_TSConv, c(1,-1), c(1,-1)) I get: Error: object R_TSConv not found If I do getNativeSymbolInfo(R_TSConv) I get: Error in FUN(R_TSConv[[1L]], ...) : no such symbol R_TSConv What am I missing? Thank you. Kevin __ 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] Foreign function call
Thank you that helps alot. Now the question is how do I know that it is in the 'stats' package? getNativeSymbolInfo doesn't seem to find 'RTSconv'. Kevin Sundar Dorai-Raj sdorai...@gmail.com wrote: You're missing that R_TSConv is an R object. You can use stats:::R_TSConv to see the value. Not sure how this helps you though. On Wed, Feb 4, 2009 at 10:08 AM, rkevinbur...@charter.net wrote: Let me get more specific. I think it this can be answered then I can translate the information to other calls. In the arima 'R' code there is a reference to .Call(R_TSconv, a, b) If from the console I type: .Call(R_TSConv, c(1,-1), c(1,-1)) I get: Error: object R_TSConv not found If I do getNativeSymbolInfo(R_TSConv) I get: Error in FUN(R_TSConv[[1L]], ...) : no such symbol R_TSConv What am I missing? Thank you. Kevin __ 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] Calling compiled functions
In the scripts I see lots of calls like .Call(.). But if I replace this exact string in my own R script I get not found. I was wondering what the scopting rules are for these functions. How just for testing (I don't want to build a full blown package) to I use .Call and .C or .Fortran foreign function calls? Thank you. Kevin __ 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
You could look into ' try' and set it up to catch errors and do the appropriate thing in your error handler. I don't have the exact syntax at hand right now but looking at ?try or ?tryCatch I think will do what you want. Kevin Alexandra Almeida alexandra...@gmail.com wrote: Hi everybody! I´m with a problem that probably is easy for you, but I really don´t know how to solve. On the following script: for(j in 1:length(limiares)) { excessos-limiares[j]-estacao[estacaolimiares[j]] par.ests-gpd(-(estacao),threshold=-limiares[j],method=c(pwm))$par.est GOF.test-ks.test(excessos,pgpd,xi=par.ests[1],beta=par.ests[2])$p.value tabs[j,]-c(par.ests,gpd(-(estacao),threshold=-limiares[j],method=c(pwm))$par.ses,GOF.test,length(excessos)) } I´ve found the error for some values of i: Erro em ks.test(excessos, pgpd, xi = par.ests[1], beta = par.ests[2]) : NA/NaN/Inf em chamada de função externa (argumento 1) *My question is: This warning stop the for and I don´t want it, is there some way to continue the for, and for the cases where the function cannot calculate the ks.test for the i just leave a NA as answer???* Thank you *very much*!!! Alexandra Almeida -- Alexandra R M de Almeida [[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] ifelse help?
I am having a hard time understanding what is happening with ifelse. Let me illustrate: h - numeric(5) p - 1:5 j - floor(j) x - 1:1000 ifelse(h == 0, x[j+2], 1:5) [1] 2 3 4 5 6 My question is, shouldn't this be retruning 25 numbers? It seems that the ifelse should check 5 values of h for zero. For each of the 5 values I am thinking it should return an array of 5 (x[j+2] if h[index] == 0). Since the dimension of h is 5 that would mean 25 values. But that isn't what is being returned.Something about this that I don't understand. Please help my ignorance. Thank you. Kevin __ 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] ifelse help?
Sorry I didn't give the proper initialization of j. But you are right j should also be an array of 5. So x[j + 5] would return 5 values. So if the array returned from 'ifelse' is the same dimention as test (h), then are all the values of h being tested? So since h as you say has no dimensions is the test only testing h[1]? Again it seems that if all of the elements of h are tested (there are 5 elements) and each element produces an array of 5 the resulting array should be 25. Kevin Charles C. Berry cbe...@tajo.ucsd.edu wrote: On Mon, 19 Jan 2009, rkevinbur...@charter.net wrote: I am having a hard time understanding what is happening with ifelse. Let me illustrate: h - numeric(5) p - 1:5 j - floor(j) And j is 0:4 + epsilon , where 0 = epsilon 1, evidently. x - 1:1000 ifelse(h == 0, x[j+2], 1:5) [1] 2 3 4 5 6 My question is, shouldn't this be retruning 25 numbers? No. It should be A vector of the same length and attributes (including class) as test and data values from the values of yes or no according to ?ifelse, and 'test' is what you have as 'h' Consider z - as.data.frame( diag(2 ) ) ifelse(z == 0, letters , 1:5) V1 V2 [1,] 1 c [2,] b 4 all args have different lengths and classes. But the result has those of the 'test' arg. It seems that the ifelse should check 5 values of h for zero. For each of the 5 values I am thinking it should return an array of 5 (x[j+2] if h[index] == 0). Since the dimension of h is 5 that would mean 25 values. h has no attributes, therefore no 'dimension' HTH, Chuck But that isn't what is being returned.Something about this that I don't understand. Please help my ignorance. Thank you. Kevin __ 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. Charles C. Berry(858) 534-2098 Dept of Family/Preventive Medicine E mailto:cbe...@tajo.ucsd.edu UC San Diego http://famprevmed.ucsd.edu/faculty/cberry/ La Jolla, San Diego 92093-0901 __ 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] Warning on assignment.
I have a question on whether a warning message is valid or if I just don't understand the process. Let me illustrate via some R code: x - 1:20 i - x %% 2 0 y - rep(1,20) x[i] - y Warning message: In x[i] - y : number of items to replace is not a multiple of replacement length But it still does what I would expect for the assignment: x [1] 1 2 1 4 1 6 1 8 1 10 1 12 1 14 1 16 1 18 1 20 What don't I understand? Thank you. Kevin __ 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] Warning on assignment.
This was just an illustration. It is the warning message that I don't understand. The warning says number of items to replace is not a multiple of replacement length. The way I look at it 10 is a multiple of 20. Kevin Sarah Goslee sarah.gos...@gmail.com wrote: The lengths are different, particularly the length of subsetted x[i] x - 1:20 i - x %% 2 0 y - rep(1,20) length(x) [1] 20 length(i) [1] 20 length(x[i]) [1] 10 length(y) [1] 20 You happened to be lucky and got what you wanted, but a more reliable approach is: x[i] - y[i] Sarah On Thu, Jan 15, 2009 at 1:08 PM, rkevinbur...@charter.net wrote: I have a question on whether a warning message is valid or if I just don't understand the process. Let me illustrate via some R code: x - 1:20 i - x %% 2 0 y - rep(1,20) x[i] - y Warning message: In x[i] - y : number of items to replace is not a multiple of replacement length But it still does what I would expect for the assignment: x [1] 1 2 1 4 1 6 1 8 1 10 1 12 1 14 1 16 1 18 1 20 What don't I understand? Thank you. Kevin -- Sarah Goslee http://www.functionaldiversity.org __ 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] Partial sort?
This is definitely a newbie question but from the documentation I have not been able to figure out what the partial sort option on the sort method does. I have read and re-read the documentation and looked at the examples but for some reason it doesn't register. Would someone attempt to explain what sort with a non-null partial array of indices does? Thank you. Kevin __ 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] pmax and sort?
I am having a hard time understanding the documentation and I was wondering if there would be someone to help clear the cobwebs. The documentation for pmax states: pmax and pmin take one or more vectors (or matrices) as arguments and return a single vector giving the ‘parallel’ maxima (or minima) of the vectors. The first element of the result is the maximum (minimum) of the first elements of all the arguments, the second element of the result is the maximum (minimum) of the second elements of all the arguments and so on. Shorter inputs are recycled if necessary. attributes (such as names or dim) are transferred from the first argument (if applicable). This descirption seems to me like what sort would return. Would someone please help me understand the differenece between pmax and sort? Thank you. Kevin __ 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] pmax and sort?
I am having a hard time understanding the documentation and I was wondering if there would be someone to help clear the cobwebs. The documentation for pmax states: pmax and pmin take one or more vectors (or matrices) as arguments and return a single vector giving the ‘parallel’ maxima (or minima) of the vectors. The first element of the result is the maximum (minimum) of the first elements of all the arguments, the second element of the result is the maximum (minimum) of the second elements of all the arguments and so on. Shorter inputs are recycled if necessary. attributes (such as names or dim) are transferred from the first argument (if applicable). This descirption seems to me like what sort would return. Would someone please help me understand the differenece between pmax and sort? Thank you. Kevin __ 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] Power functions?
I had a question about the basic power functions in R. For example from the R console I enter: -1 ^ 2 [1] -1 but also -1^3 [1] -1 -0.1^2 [1] -0.01 Normally pow(-1, 2) return either -Infinity or NaN. Has R taken over the math functions? If so I would think that -1^2 is 1 not -1 and -0.1^2 is 0.01 not -0.01. Thank you. Kevin __ 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.