Re: [R] data.table error
Johannes, please try the latest version on R-forge (1.4). That error has been fixed, and it's much faster. We hope to have that to CRAN reasonably soon. To install, use: install.packages("data.table",repos="http://R-Forge.R-project.org";) - Tom Tom Short On Thu, Apr 29, 2010 at 3:40 PM, johannes rara wrote: > I'm trying to learn data.table package but I get a following annoying > error message: > >> install.packages("data.table") > trying URL > 'http://www.freestatistics.org/cran/bin/macosx/universal/contrib/2.10/data.table_1.2.tgz' > Content type 'application/x-gzip' length 66823 bytes (65 Kb) > opened URL > == > downloaded 65 Kb > > > The downloaded packages are in > > /var/folders/n-/n-wPTanPGa4PpVd0bTgCOU+++TI/-Tmp-//RtmppqPptG/downloaded_packages >> library(data.table) >> cr <- data.table(cars) >> cr[speed == 20] > speed dist > [1,] 20 32 > [2,] 20 48 > [3,] 20 52 > [4,] 20 56 > [5,] 20 64 > Warning messages: > 1: In `[.data.table`(cr, speed == 20) : > This R session is < 2.4.0. Please upgrade to 2.4.0+. > 2: In `[.data.table`(cr, speed == 20) : > This R session is < 2.4.0. Please upgrade to 2.4.0+. >> > > I'm using R 2.10.1 (see sessioninfo below), so why this error message > keeps popping up? > >> sessionInfo() > R version 2.10.1 (2009-12-14) > i386-apple-darwin8.11.1 > > locale: > [1] fi_FI.UTF-8/fi_FI.UTF-8/C/C/fi_FI.UTF-8/fi_FI.UTF-8 > > attached base packages: > [1] stats graphics grDevices utils datasets methods base > > other attached packages: > [1] data.table_1.2 ref_0.97 > > loaded via a namespace (and not attached): > [1] tools_2.10.1 >> > > -J > > __ > 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] Code is too slow: mean-centering variables in a data framebysubgroup
Another way that Matthew Dowle showed me for this type of problem is to reshape frame to a long format. It makes it easier to manipulate and can be faster. > longdt <- with(frame, data.table(group = unlist(rep(group, each=7)), x = > c(a,b,c,d,e,f,g))) > > system.time(new.frame4 <- longdt[, x/mean(x, na.rm = TRUE), by = "group"]) user system elapsed 0.540.040.61 > > # Or, remove the NAs ahead of time for more speed: > > longdt2 <- longdt[!is.na(longdt$x),] > system.time(new.frame4 <- longdt2[, x/mean(x), by = "group"]) user system elapsed 0.17 0.000.17 - Tom On Wed, Apr 7, 2010 at 3:46 PM, Tom Short wrote: > Here's how I would have done the data.table method. It's a bit faster > than the ave approach on my machine: > >> # install.packages("data.table",repos="http://R-Forge.R-project.org";) >> library(data.table) >> >> f3 <- function(frame) { > + frame <- as.data.table(frame) > + frame[, lapply(.SD[,2:ncol(.SD), with = FALSE], > + function(x) x / mean(x, na.rm = TRUE)), > + by = "group"] > + } >> >> system.time(new.frame2 <- f2(frame)) # ave > user system elapsed > 0.50 0.08 1.24 >> system.time(new.frame3 <- f3(frame)) # data.table > user system elapsed > 0.25 0.01 0.30 > > - Tom > > Tom Short > > > On Wed, Apr 7, 2010 at 12:46 PM, Dimitri Liakhovitski > wrote: >> I would like to thank once more everyone who helped me with this question. >> I compared the speed for different approaches. Below are the results >> of my comparisons - in case anyone is interested: >> >> ### Building an EXAMPLE FRAME with N rows - with groups and a lot of NAs: >> N<-10 >> set.seed(1234) >> frame<-data.frame(group=rep(paste("group",1:10),N/10),a=rnorm(1:N),b=rnorm(1:N),c=rnorm(1:N),d=rnorm(1:N),e=rnorm(1:N),f=rnorm(1:N),g=rnorm(1:N)) >> frame<-frame[order(frame$group),] >> >> ## Introducing 60% NAs: >> names.used<-names(frame)[2:length(frame)] >> set.seed(1234) >> for(i in names.used){ >> i.for.NA<-sample(1:N,round((N*.6),0)) >> frame[[i]][i.for.NA]<-NA >> } >> lapply(frame[2:8], function(x) length(x[is.na(x)])) # Checking that it worked >> ORIGframe<-frame ## placeholder for the unchanged original frame >> >> ### Objective of the code - divide each value by its group mean >> >> ### METHOD 1 - the FASTEST - using ave():## >> frame<-ORIGframe >> f2 <- function(frame) { >> for(i in 2:ncol(frame)) { >> frame[,i] <- ave(frame[,i], frame[,1], >> FUN=function(x)x/mean(x,na.rm=TRUE)) >> } >> frame >> } >> system.time({new.frame<-f2(frame)}) >> # Took me 0.23-0.27 sec >> ### >> >> ### METHOD 2 - fast, just a bit slower - using data.table: >> ## >> >> # If you don't have it - install the package - NOT from CRAN: >> install.packages("data.table",repos="http://R-Forge.R-project.org";) >> library(data.table) >> frame<-ORIGframe >> system.time({ >> table<-data.table(frame) >> colMeanFunction<-function(data,key){ >> data[[key]]=NULL >> ret=as.matrix(data)/matrix(rep(as.numeric(colMeans(as.data.frame(data),na.rm=T)),nrow(data)),nrow=nrow(data),ncol=ncol(data),byrow=T) >> return(ret) >> } >> groupedMeans = table[,colMeanFunction(.SD, "group"), by="group"] >> names.to.use<-names(groupedMeans) >> for(i in >> 1:length(groupedMeans)){groupedMeans[[i]]<-as.data.frame(groupedMeans[[i]])} >> groupedMeans<-do.call(cbind, groupedMeans) >> names(groupedMeans)<-names.to.use >> }) >> # Took me 0.37-.45 sec >> ### >> >> ### METHOD 3 - fast, a tad slower (using model.matrix & matrix >> multiplication):## >> frame<-ORIGframe >> system.time({ >> mat <- as.matrix(frame[,-1]) >> mm <- model.matrix(~0+group,frame) >> col.grp.N <- crossprod( !is.na(mat), mm ) # Use this line if don't >> want to use NAs for mean calculations >> # col.grp.N <- crossprod( mat != 0 , mm ) # Use this line if don't >> want to use zeros for mean calculations >> mat[is.na(mat)] <- 0.0 >> col.grp.sum <- crossprod( mat, mm ) >> mat <- mat / ( t(col.grp.sum/col.grp.N)[ frame$group,] ) >> is.na(mat) <- is.na(frame[,-1]) >&g
Re: [R] Code is too slow: mean-centering variables in a data framebysubgroup
Here's how I would have done the data.table method. It's a bit faster than the ave approach on my machine: > # install.packages("data.table",repos="http://R-Forge.R-project.org";) > library(data.table) > > f3 <- function(frame) { + frame <- as.data.table(frame) + frame[, lapply(.SD[,2:ncol(.SD), with = FALSE], + function(x) x / mean(x, na.rm = TRUE)), + by = "group"] + } > > system.time(new.frame2 <- f2(frame)) # ave user system elapsed 0.500.081.24 > system.time(new.frame3 <- f3(frame)) # data.table user system elapsed 0.250.010.30 - Tom Tom Short On Wed, Apr 7, 2010 at 12:46 PM, Dimitri Liakhovitski wrote: > I would like to thank once more everyone who helped me with this question. > I compared the speed for different approaches. Below are the results > of my comparisons - in case anyone is interested: > > ### Building an EXAMPLE FRAME with N rows - with groups and a lot of NAs: > N<-10 > set.seed(1234) > frame<-data.frame(group=rep(paste("group",1:10),N/10),a=rnorm(1:N),b=rnorm(1:N),c=rnorm(1:N),d=rnorm(1:N),e=rnorm(1:N),f=rnorm(1:N),g=rnorm(1:N)) > frame<-frame[order(frame$group),] > > ## Introducing 60% NAs: > names.used<-names(frame)[2:length(frame)] > set.seed(1234) > for(i in names.used){ > i.for.NA<-sample(1:N,round((N*.6),0)) > frame[[i]][i.for.NA]<-NA > } > lapply(frame[2:8], function(x) length(x[is.na(x)])) # Checking that it worked > ORIGframe<-frame ## placeholder for the unchanged original frame > > ### Objective of the code - divide each value by its group mean > > ### METHOD 1 - the FASTEST - using ave():## > frame<-ORIGframe > f2 <- function(frame) { > for(i in 2:ncol(frame)) { > frame[,i] <- ave(frame[,i], frame[,1], > FUN=function(x)x/mean(x,na.rm=TRUE)) > } > frame > } > system.time({new.frame<-f2(frame)}) > # Took me 0.23-0.27 sec > ### > > ### METHOD 2 - fast, just a bit slower - using data.table: > ## > > # If you don't have it - install the package - NOT from CRAN: > install.packages("data.table",repos="http://R-Forge.R-project.org";) > library(data.table) > frame<-ORIGframe > system.time({ > table<-data.table(frame) > colMeanFunction<-function(data,key){ > data[[key]]=NULL > ret=as.matrix(data)/matrix(rep(as.numeric(colMeans(as.data.frame(data),na.rm=T)),nrow(data)),nrow=nrow(data),ncol=ncol(data),byrow=T) > return(ret) > } > groupedMeans = table[,colMeanFunction(.SD, "group"), by="group"] > names.to.use<-names(groupedMeans) > for(i in > 1:length(groupedMeans)){groupedMeans[[i]]<-as.data.frame(groupedMeans[[i]])} > groupedMeans<-do.call(cbind, groupedMeans) > names(groupedMeans)<-names.to.use > }) > # Took me 0.37-.45 sec > ### > > ### METHOD 3 - fast, a tad slower (using model.matrix & matrix > multiplication):## > frame<-ORIGframe > system.time({ > mat <- as.matrix(frame[,-1]) > mm <- model.matrix(~0+group,frame) > col.grp.N <- crossprod( !is.na(mat), mm ) # Use this line if don't > want to use NAs for mean calculations > # col.grp.N <- crossprod( mat != 0 , mm ) # Use this line if don't > want to use zeros for mean calculations > mat[is.na(mat)] <- 0.0 > col.grp.sum <- crossprod( mat, mm ) > mat <- mat / ( t(col.grp.sum/col.grp.N)[ frame$group,] ) > is.na(mat) <- is.na(frame[,-1]) > mat<-as.data.frame(mat) > }) > # Took me 0.44-0.50 sec > ### > > ### METHOD 5- much slower - it's the one I started > with:## > frame<-ORIGframe > system.time({ > frame <- do.call(cbind, lapply(names.used, function(x){ > unlist(by(frame, frame$group, function(y) y[,x] / mean(y[,x],na.rm=T))) > })) > }) > # Took me 1.25-1.32 min > ### > > ### METHOD 6 - the slowest; using "plyr" and > "ddply":## > frame<-ORIGframe > library(plyr) > function3 <- function(x) x / mean(x, na.rm = TRUE) > system.time({ > grouping.factor<-"group" > myvariables<-names(frame)[2:8] > frame3<-ddply(frame, grouping.factor, colwise(function3, myvariables)) > }) > # Took me 1.36-1.47 min > ### > > > Thanks again! > Dimitri > > > On Wed, Mar 31, 2010 at 8:29 PM, William Dunlap wrote: >> Dimitri,
Re: [R] rpad ?
As the author of Rpad, I'll say that it is officially abandoned. I just don't have the time or the need for my job. If someone is interested in maintaining it, I'll try to answer questions (the email address listed on the package hasn't worked for a while, and the mailing list got overwhelmed with spam). Of the other R web interfaces I've played with or looked at, RApache is the most promising. It offers more performance and security than the Rpad approach. You can also make some pretty interactive pages. The trade-off is that it's harder to build applications (the last time I looked anyway). To get interactivity, the RApache approach requires a fair amount of javascript programming. Rpad gives you interactivity fairly automatically as a webpage with embedded R code. - Tom Tom Short On Tue, Mar 23, 2010 at 4:46 PM, Erich Neuwirth wrote: > We are using RPad for a teaching application here. > But we had to find many things the hard way, > and additionally, it did not survive the latest R release change. > There is a minimal repair, but the maintainer does not answer any email > any more. We did the repair and are giving a modified version to our > students, but we do not have enough resource to take over maintenance. > > > > On 3/23/2010 8:00 PM, sjaffe wrote: >> >> Is anyone using rpad? Is there any documentation or examples beyond that in >> the 'man' directory of the source? >> > > -- > Erich Neuwirth, University of Vienna > Faculty of Computer Science > Computer Supported Didactics Working Group > Visit our SunSITE at http://sunsite.univie.ac.at > Phone: +43-1-4277-39464 Fax: +43-1-4277-39459 > > __ > 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] data.table evaluating columns
On Tue, Mar 2, 2010 at 7:09 PM, Rob Forler wrote: > Hi everyone, > > I have the following code that works in data frames taht I would like tow > ork in data.tables . However, I'm not really sure how to go about it. > > I basically have the following > > names = c("data1", "data2") > frame = data.frame(list(key1=as.integer(c(1,2,3,4,5,6)), > key2=as.integer(c(1,2,3,2,5,6)),data1 = c(3,3,2,3,5,2), data2= > c(3,3,2,3,5,2))) > > for(i in 1:length(names)){ > frame[, paste(names[i], "flag")] = frame[,names[i]] < 3 > > } > > Now I try with data.table code: > names = c("data1", "data2") > frame = data.table(list(key1=as.integer(c(1,2,3,4,5,6)), > key2=as.integer(c(1,2,3,2,5,6)),data1 = c(3,3,2,3,5,2), data2= > c(3,3,2,3,5,2))) > > for(i in 1:length(names)){ > frame[, paste(names[i], "flag"), with=F] = as.matrix(frame[,names[i], > with=F] )< 3 > > } Rob, this type of question is better for the package maintainer(s) directly rather than R-help. That said, one answer is to use list addressing: for(i in 1:length(names)){ frame[[paste(names[i], "flag")]] = frame[[names[i]]] < 3 } Another option is to manipulate frame as a data frame and convert to data.table when you need that functionality (conversion is quick). In the data table version, frame[,names[i], with=F] is the same as frame[,names[i], drop=FALSE] (the answer is a list, not a vector). Normally, it's easier to use [[]] or $ indexing to get this. Also, fname[i,j] <- something assignment is still a bit buggy for data.tables. - Tom Tom Short __ 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] dramatic speed difference in lapply
I'm sorry, Rob, but that code is dense enough and formatted badly enough that it's hard to dig through. You may want to try the data.table package. The development version on R-forge is pretty fast for grouping operations like this. I'm not sure if this is what you're really after. It's hard to tell from your example. Compare some speeds: > dat <- data.frame(D=sample(32000:33000, 666000,T), + Fid=sample(1:10,666000,T), + A=sample(1:5,666000,T)) > > ### one of your examples > system.time(ret <- fedb.ddplyWrapper2(dat, c("D", "Fid"), + function(x) c(sum(x[,"A"], na.rm=T), sum(x[,"A"], na.rm=T user system elapsed 21.78 14.42 36.35 > > > ### data.table > install.packages("data.table",repos="http://R-Forge.R-project.org";) > library(data.table) > dt <- as.data.table(dat) > system.time(ret2 <- dt[, sum(A, na.rm=T), by = "D,Fid"]) user system elapsed 0.270.000.28 > > > ### plyr for comparison, too > library(plyr) > system.time(ret3 <- ddply(dat, .(D,Fid), function(x) sum(x$A, na.rm=T))) user system elapsed 28.94 12.16 41.23 > head(ret) [,1] [,2] 1 175 175 2 222 222 3 221 221 4 134 134 5 253 253 6 194 194 > head(ret2) D Fid V1 [1,] 32000 1 228 [2,] 32000 2 209 [3,] 32000 3 182 [4,] 32000 4 180 [5,] 32000 5 181 [6,] 32000 6 222 > head(ret3) D Fid V1 1 32000 1 175 2 32000 2 222 3 32000 3 221 4 32000 4 134 5 32000 5 253 6 32000 6 194 - Tom On Fri, Feb 26, 2010 at 2:58 PM, Rob Forler wrote: > So I have a function that does lapply's for me based on dimension. Currently > only works for length(pivotColumns)=2 because I haven't fixed the rbinds. I > have two versions. One runs WAYYY faster than the other. And I'm not sure > why. > > Fast Version: > > fedb.ddplyWrapper2Fast <- function(data, pivotColumns, listNameFunctions, > ...){ > lapplyFunctionRecurse <- function(cdata, level=1, ...){ > if(level==1){ > > return(lapply(split(seq(nrow(cdata)),cdata[,pivotColumns[level]], drop=T), > function(x) lapplyFunctionRecurse(x, level+1, ...))) > } else if (level==length(pivotColumns)) { > # > return(lapply(split(cdata,data[cdata,pivotColumns[level]], drop=T), > function(x, ...) listNameFunctions(data[x,], ...))) > return(lapply(split(cdata,data[cdata,pivotColumns[level]], > drop=T), function(x, ...) c(data[cdata[1],pivotColumns[2]], > data[cdata[1],pivotColumns[1]], sum(data[cdata,"A"], na.rm=T), > sum(data[cdata,"A"], na.rm=T > } else { > return(lapply(split(cdata,data[cdata,pivotColumns[level]], > drop=T), function(x) lapplyFunctionRecurse(x, level+1, ...))) > } > } > result = lapplyFunctionRecurse(data, ...) > matrix2 <- do.call('rbind', lapply(result, function(x) > do.call('rbind',x))) > return(matrix2) > } > > > dat <- data.frame(D=sample(32000:33000, 666000, > T),Fid=sample(1:10,666000,T), A=sample(1:5,666000,T)) >> temp = proc.time(); ret = fedb.ddplyWrapper2(dat, c("D", "Fid"), > function(x) c(sum(x[,"A"], na.rm=T), sum(x[,"A"], na.rm=T))); > proc.time()-temp > user system elapsed > 4.616 0.006 4.630 > #note in thie case the anonymous function I pass in isn't used because I > hardcode the function into the lapply. > > approx 4 seconds > > This runs very fast. This runs very slow: > > fedb.ddplyWrapper2 <- function(data, pivotColumns, listNameFunctions, ...){ > lapplyFunctionRecurse <- function(cdata, level=1, ...){ > if(level==1){ > > return(lapply(split(seq(nrow(cdata)),cdata[,pivotColumns[level]], drop=T), > function(x) lapplyFunctionRecurse(x, level+1, ...))) > } else if (level==length(pivotColumns)) { > #this line is different. it essentially calls the function you > pass in > return(lapply(split(cdata,data[cdata,pivotColumns[level]], > drop=T), function(x, ...) listNameFunctions(data[x,], ...))) > } else { > return(lapply(split(cdata,data[cdata,pivotColumns[level]], > drop=T), function(x) lapplyFunctionRecurse(x, level+1, ...))) > } > } > result = lapplyFunctionRecurse(data, ...) > matrix2 <- do.call('rbind', lapply(result, function(x) > do.call('rbind',x))) > return(matrix2) > } > > dat <- data.frame(D=sample(32000:33000, 666000, > T),Fid=sample(1:10,666000,T), A=sample(1:5,666000,T)) >> temp = proc.time(); ret = fedb.ddplyWrapper2(dat, c("D", "Fid"), > function(x) c(sum(x[,"A"], na.rm=T), sum(x[,"A"], na.rm=T))); > proc.time()-temp > user system elapsed > 16.346 65.059 81.680 > > head(ret3) D Fid V1 1 32000 1 175 2 32000 2 222 3 32000 3 221 4 32000 4 134 5 32000 5 253 6 32000 6 194 > > > Can anyone explain to me why there is a 4x time difference? I don't want to > have to hardcore into the recursion function, but if I have to I will. > > Thanks, > Rob > > [[alternative HTML version deleted]] > > __ > R-help@r-
Re: [R] how to fast extract values from different list elements
On Thu, Feb 25, 2010 at 4:10 AM, Heym, Peter-Paul wrote: > this works fine but it is very slow (since A and B can be very large and I > have to repeat this about 5000 times). I would like to make this faster using > e.g. apply or lapply but I didn't get it work using these methods. Does > anybody know an EFFICIENT or FAST way extract the values from L using the > values from A and B? Instead of L[[A[i]]][B[i]], try L[A][B] - Tom __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] how to rearrange a dataframe
Try this: a <- b <- read.table(textConnection(" 1 + name1 1 2 3 2 + name2 5 9 10 2 - name3 56 74 93 1 - name4 65 75 98"), skip=1, header=FALSE) swapidx <- with(a, (V1 == 2 & V2 == "+") | (V1 == 1 & V2 == "-")) b[swapidx,] <- b[swapidx, c(1:3,6:4)] This creates an indexing vector that identifies which rows to swap, then the 6:4 flips around the fourth through sixth columns. - Tom On Tue, Feb 23, 2010 at 5:27 PM, Laura Rodriguez Murillo wrote: > Hi all, > > I'd appreciate if anyone can help me with this... > > I have a data frame that looks like this: > > 1 + name1 1 2 3 > 2 + name2 5 9 10 > 2 - name3 56 74 93 > 1 - name4 65 75 98 > > I need to rearrange this in a way so that the rows with "1" in the > first column, and "-" in the second column; then columns 4 and 6 > should switch places. That is, column 6 would be now column 4 and > column 4 would be column 6 (column 5 should stay as column 5) > In the same way, if the first column is "2" and the second is "+", > then the same rearrangement should be done. > Rows with the first two entries 1 + or 2 - should stay in the same order. > This should be done for each row independently. > > Thanks a lot for your help! > > __ > 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] Large dataset importing, columns merging and splitting
If you need more aggregations on the stock (I assume that's what the first column is), I'd use the data.table package. It allows fast indexing and merge operations. That's handy if you have other features of a stock (like company size or industry sector) that you'd like to include in the aggregation. Like Gabor, I'd probably use chron for keeping track of the dates. Here's some code to get you started: Lines <- "CVX 20070201 9 30 51 73.25 81400 0 CVX 20070201 9 30 51 73.25 100 0 CVX 20070201 9 30 51 73.25 100 0 CVX 20070201 9 30 51 73.25 300 0 CVX 20070201 9 30 51 73.25 81400 0 CVX 20070201 9 40 51 74.25 100 0 CVX 20070201 9 40 52 74.25 100 0 CVX 20070201 9 40 53 74.25 300 0 CVX 20070301 9 30 51 74.25 100 0 CVX 20070301 9 30 51 74.25 100 0 CVX 20070301 9 30 51 74.25 300 0 CVX 20070301 9 30 51 74.25 81400 0 CVX 20070301 9 40 51 74.25 100 0 CVX 20070301 9 40 52 74.25 100 0 CVX 20070301 9 40 53 74.25 300 0 DVX 20070201 9 30 51 73.25 81400 0 DVX 20070201 9 30 51 73.25 100 0 DVX 20070201 9 30 51 73.25 100 0 DVX 20070201 9 30 51 73.25 300 0 DVX 20070201 9 30 51 73.25 81400 0 DVX 20070201 9 40 51 74.25 100 0 DVX 20070201 9 40 52 74.25 100 0 DVX 20070201 9 40 53 74.25 300 0 DVX 20070301 9 30 51 74.25 100 0 DVX 20070301 9 30 51 74.25 100 0 DVX 20070301 9 30 51 74.25 300 0 DVX 20070301 9 30 51 74.25 81400 0 DVX 20070301 9 40 51 74.25 100 0 DVX 20070301 9 40 52 74.25 100 0 DVX 20070301 9 40 53 74.25 300 0" library(data.table) library(chron) dt <- data.table(read.table(textConnection(Lines), colClasses = c("character", "numeric", "numeric", "numeric", "numeric", "numeric", "numeric", "numeric"), col.names = c("stock", "date", "h", "m", "s", "Price", "Volume", "xx"))) dt$date <- as.chron(as.Date(as.character(dt$date), format = "%Y%m%d")) + dt$h/24 + dt$m/(60*24) + dt$s/(60*60*24) dt$roundeddate <- as.integer(floor(as.numeric(dt$date) * (24 * 12))) # data.table likes integers dt[,list(meanprice = mean(Price), volume = sum(Volume)), by = "roundeddate"] dt[,list(meanprice = mean(Price), volume = sum(Volume)), by = "stock,roundeddate"] You'd still probably want to turn the roundeddate back into a real chron object. If you use aggregation a lot, the development version of data.table has faster aggregations: http://r-forge.r-project.org/projects/datatable/ - Tom On Tue, Jan 26, 2010 at 11:23 AM, Gabor Grothendieck wrote: > Try this using the development version of read.zoo in zoo (which we > source from the R-Forge on the fly). > > We use "NULL" in colClasses for those columns we don't need but in > col.names we still have to include dummy names for > them. Of what is left the index is the first three columns (1:3) > which we convert to chron class times in FUN and then truncate to 5 > seconds in FUN2. Finally we use aggregate = mean to average over the > 5 second intervals. > > Lines <- "CVX 20070201 9 30 51 73.25 81400 0 > CVX 20070201 9 30 51 73.25 100 0 > CVX 20070201 9 30 51 73.25 100 0 > CVX 20070201 9 30 51 73.25 300 0 > CVX 20070201 9 30 51 73.25 81400 0 > CVX 20070201 9 40 51 73.25 100 0 > CVX 20070201 9 40 52 73.25 100 0 > CVX 20070201 9 40 53 73.25 300 0" > > > library(zoo) > source("http://r-forge.r-project.org/plugins/scmsvn/viewcvs.php/*checkout*/pkg/zoo/R/read.zoo.R?rev=611&root=zoo";) > library(chron) > > z <- read.zoo(textConnection(Lines), > colClasses = c("NULL", "NULL", "numeric", "numeric", "numeric", > "numeric", > "numeric", "NULL"), > col.names = c("V1", "V2", "V3", "V4", "V5", "Price", "Volume", "V8"), > index = 1:3, > FUN = function(tt) times(paste(tt[,1], tt[,2], tt[,3], sep = ":")), > FUN2 = function(tt) trunc(tt, "00:00:05"), > aggregate = mean) > > The result of running the above is: > >> z > Price Volume > 09:30:50 73.25 32660. > 09:40:50 73.25 166.6667 > > On Tue, Jan 26, 2010 at 10:48 AM, Manta wrote: >> >> Dear All, >> I have a large data set that looks like this: >> >> CVX 20070201 9 30 51 73.25 81400 0 >> CVX 20070201 9 30 51 73.25 100 0 >> CVX 20070201 9 30 51 73.25 100 0 >> CVX 20070201 9 30 51 73.25 300 0 >> >> First, I would like to import it by merging column 3 4 and 5, since that is >> the timestamp. Then, I would like to aggregate the data by splitting them in >> bins of 5 minutes size, therefore from 93000 up to 93459 etc, givin as >> output the average price and volume in the 5 minutes bin. >> >> Hope this helps, >> Best, >> >> Marco >> -- >> View this message in context: >> http://n4.nabble.com/Large-dataset-importing-columns-merging-and-splitting-tp1294668p1294668.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. >
Re: [R] Getting file name from pdf device?
On Fri, Jul 31, 2009 at 8:49 AM, Rainer M Krug wrote: > My question: how can I get the filename of the pdf from the device > before it is closed? I've also looked for this and couldn't find a way. I had a similar use, where I wanted to get an R transcript with embedded plots in emacs (see prettyR for another transcript-with-plots option). What I did was use dev2bitmap to write out a PNG file. You could do something similar with dev.copy2pdf to create the pdf after you do the plotting. You could also use dev2bitmap in this manner to drive ghostscript to create pdf's for you (I don't know if it'll compress like you want). Here's what I did: show <- function(file = paste(tempfile(), ".png", sep = "")) { dev2bitmap(file) cat("[[", file, "]]\n", sep = "") # I do some post-processing in emacs to see the embedded graphic } My use case was that plots would be inserted where I used "show" as follows: plot(sin) show()# < plot inserted into transcript here plot(cos) show("cos.png") # this time, a named local file instead of a temp file - Tom __ 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] Excel Export in a beauty way
Another useful way to create a formatted Excel file is to write out an HTML file, but put an XLS extension on it. When Excel reads it, it will convert it. Users will treat it like an Excel file. This trick allows you to add formatted titles, table footnotes, links to other files (pdf graphs for example), and more. To create HTML, you have several packages that can help you out: R2HTML, Rpad, hwriter, and xtable. Not everything might convert properly, so you may have to experiment. Data frames as tables normally convert nicely. - Tom Tom Short __ 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] Do you use R for data manipulation?
Another tool I find useful is Matthew Dowle's data.table package. It has very fast indexing, can have much lower memory requirements than a data frame, and has some built-in data manipulation capability. Especially with a 64-bit OS, you can use this to keep things in memory where you otherwise would have to use a database. See here: http://article.gmane.org/gmane.comp.lang.r.packages/282 - Tom __ 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.