Re: [R] expand a matrix
You can specify multiple indexes to replace at once, so you can avoid a for loop entirely like this: M <- matrix(nrow = 10, ncol = 10) M[1:4, 5: 6] <- M[5: 6, 1:4] <- m[1, 2] M[1:4,7] <- M[ 7, 1:4] <- m[1, 3] M[1:4, 8:10] <- M[8:10, 1:4] <- m[1, 4] M[5:6,7] <- M[ 7, 5:6] <- m[2, 3] M[5:6, 8:10] <- M[8:10, 5:6] <- m[2, 4] M[ 7, 8:10] <- M[8:10, 7] <- m[3, 4] M You could also specify it directly with function 'matrix', it is somewhat ugly though: M <- matrix(c( NA, NA, NA, NA, m[1,2], m[1,2], m[1,3], m[1,4], m[1,4], m[1,4], NA, NA, NA, NA, m[1,2], m[1,2], m[1,3], m[1,4], m[1,4], m[1,4], NA, NA, NA, NA, m[1,2], m[1,2], m[1,3], m[1,4], m[1,4], m[1,4], NA, NA, NA, NA, m[1,2], m[1,2], m[1,3], m[1,4], m[1,4], m[1,4], m[1,2], m[1,2], m[1,2], m[1,2], NA, NA, m[2,3], m[2,4], m[2,4], m[2,4], m[1,2], m[1,2], m[1,2], m[1,2], NA, NA, m[2,3], m[2,4], m[2,4], m[2,4], m[1,3], m[1,3], m[1,3], m[1,3], m[2,3], m[2,3], NA, m[3,4], m[3,4], m[3,4], m[1,4], m[1,4], m[1,4], m[1,4], m[2,4], m[2,4], m[3,4], NA, NA, NA, m[1,4], m[1,4], m[1,4], m[1,4], m[2,4], m[2,4], m[3,4], NA, NA, NA, m[1,4], m[1,4], m[1,4], m[1,4], m[2,4], m[2,4], m[3,4], NA, NA, NA ), nrow = 10, ncol = 10, byrow = TRUE) M If it interests you, you could take a look at package 'Matrix', it contains classes for specific types of matrices, including upper triangular matrices. This package helps to improve speed of linear algebra computations and reduce the memory size of matrices. On Mon, Sep 5, 2022 at 10:46 PM Jinsong Zhao wrote: > > Hi there, > > I have a matrix likes: > > m > [,1] [,2] [,3] [,4] > [1,] NA123 > [2,]1 NA65 > [3,]26 NA4 > [4,]354 NA > > I hope to expand it to 10 by 10 matrix, M, likes: > > M >[,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] > [1,] NA NA NA NA11233 3 > [2,] NA NA NA NA11233 3 > [3,] NA NA NA NA11233 3 > [4,] NA NA NA NA11233 3 > [5,]1111 NA NA655 5 > [6,]1111 NA NA655 5 > [7,]222266 NA44 4 > [8,]3333554 NA NANA > [9,]3333554 NA NANA > [10,]3333554 NA NANA > > I use the following code: > > M <- matrix(NA, 10, 10) > > for (i in 1:10) { > for (j in 1:10) { >if (i %in% 1:4 & j %in% 5:6) M[i,j] <- M[j,i] <- m[1,2] >if (i %in% 1:4 & j == 7) M[i,j] <- M[j,i] <-m[1,3] >if (i %in% 1:4 & j %in% 8:10) M[i,j] <- M[j,i] <-m[1,4] >if (i %in% 5:6 & j == 7) M[i,j] <- M[j,i] <-m[2,3] >if (i %in% 5:6 & j %in% 8:10) M[i,j] <- M[j,i] <-m[2,4] >if (i == 7 & j %in% 8:10) M[i,j] <- M[j,i] <-m[3,4] > } > } > > Is there any convenience way to do it? Thanks! > > Best, > Jinsong > > __ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > 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 -- To UNSUBSCRIBE and more, see 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] expand a matrix
Hi there, I have a matrix likes: > m [,1] [,2] [,3] [,4] [1,] NA123 [2,]1 NA65 [3,]26 NA4 [4,]354 NA I hope to expand it to 10 by 10 matrix, M, likes: > M [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [1,] NA NA NA NA11233 3 [2,] NA NA NA NA11233 3 [3,] NA NA NA NA11233 3 [4,] NA NA NA NA11233 3 [5,]1111 NA NA655 5 [6,]1111 NA NA655 5 [7,]222266 NA44 4 [8,]3333554 NA NANA [9,]3333554 NA NANA [10,]3333554 NA NANA I use the following code: M <- matrix(NA, 10, 10) for (i in 1:10) { for (j in 1:10) { if (i %in% 1:4 & j %in% 5:6) M[i,j] <- M[j,i] <- m[1,2] if (i %in% 1:4 & j == 7) M[i,j] <- M[j,i] <-m[1,3] if (i %in% 1:4 & j %in% 8:10) M[i,j] <- M[j,i] <-m[1,4] if (i %in% 5:6 & j == 7) M[i,j] <- M[j,i] <-m[2,3] if (i %in% 5:6 & j %in% 8:10) M[i,j] <- M[j,i] <-m[2,4] if (i == 7 & j %in% 8:10) M[i,j] <- M[j,i] <-m[3,4] } } Is there any convenience way to do it? Thanks! Best, Jinsong __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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] expand gridded matrix to higher resolution
Hi Jeff, thanks, the raster package disaggregate will do the trick as well. library(raster) rmm <- raster(ncols=5, nrows=3) rmm[] <- matrix(1:15,nrow=3,byrow = T) xrmm <- disaggregate(rmm, fact=c(3, 3)) > > as.matrix(rmm) > [,1] [,2] [,3] [,4] [,5] > [1,]12345 > [2,]6789 10 > [3,] 11 12 13 14 15 > > as.matrix(xrmm) > [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13] > [,14] [,15] > [1,]111222333 4 4 4 5 > 5 5 > [2,]111222333 4 4 4 5 > 5 5 > [3,]111222333 4 4 4 5 > 5 5 > [4,]666777888 9 9 910 > 1010 > [5,]666777888 9 9 910 > 1010 > [6,]666777888 9 9 910 > 1010 > [7,] 11 11 11 12 12 12 13 13 1314141415 > 1515 > [8,] 11 11 11 12 12 12 13 13 1314141415 > 1515 > [9,] 11 11 11 12 12 12 13 13 1314141415 > 1515 the disaggregate as a bit faster than the tapply. mmb=matrix(1:259200,nrow=720,ncol=360) rmmb <- raster(ncols=360, nrows=720) rmmb[] <- mmb[] system.time(for(i in 1:10) {xmm=matrix(NA,nrow=nrow(mmb)*3,ncol=ncol(mmb)*3) for(icol in 1:ncol(mmb)) { for(irow in 1:nrow(mmb)) { xicol=(icol-1)*3 +c(1:3) xirow=(irow-1)*3 +c(1:3) xmm[xirow,xicol]=mmb[irow,icol] } } }) system.time(for(i in 1:10) {apply(t(apply(mmb,1,rep,each=3)),2,rep,each=3)}) #ca. 10x faster system.time(for(i in 1:10) {xrmmb <- disaggregate(rmmb, fact=c(3, 3))}) > > system.time(for(i in 1:10) {xmm=matrix(NA,nrow=nrow(mmb)*3,ncol=ncol(mmb)*3) > + for(icol in 1:ncol(mmb)) { > + for(irow in 1:nrow(mmb)) { > + xicol=(icol-1)*3 +c(1:3) > + xirow=(irow-1)*3 +c(1:3) > + xmm[xirow,xicol]=mmb[irow,icol] > + } > + } > + }) >user system elapsed > 8.297 0.048 8.364 > > system.time(for(i in 1:10) > > {apply(t(apply(mmb,1,rep,each=3)),2,rep,each=3)}) #ca. 10x faster >user system elapsed > 0.785 0.093 0.881 > > system.time(for(i in 1:10) {xrmmb <- disaggregate(rmmb, fact=c(3, 3))}) >user system elapsed > 0.583 0.147 0.731 cheers Peter > On 5. Jul 2017, at 16:57, Jeff Newmillerwrote: > > You probably ought to be using the raster package. See the CRAN Spatial Task > View. > -- > Sent from my phone. Please excuse my brevity. > > On July 5, 2017 12:20:28 AM PDT, "Anthoni, Peter (IMK)" > wrote: >> Hi all, >> (if me email goes out as html, than my email client don't do as told, >> and I apologies already.) >> >> We need to downscale climate data and therefore first need to expand >> the climate from 0.5deg to the higher resolution 10min, before we can >> add high resolution deviations. We basically need to have the original >> data at each gridcell replicated into 3x3 gridcells. >> A simple for loop can do this, but I could need a faster procedure. >> Anybody know a faster way? Is there package than can do what we need >> already? >> I tried matrix with rep, but I am missing some magic there, since it >> doesn't do what we need. >> replicate might be promising, but then still need to rearrange the >> output into the column and row format we need. >> >> A simple example: >> mm=matrix(1:15,nrow=3,byrow = T) >> xmm=matrix(NA,nrow=nrow(mm)*3,ncol=ncol(mm)*3) >> for(icol in 1:ncol(mm)) { >> for(irow in 1:nrow(mm)) { >> xicol=(icol-1)*3 +c(1:3) >> xirow=(irow-1)*3 +c(1:3) >> xmm[xirow,xicol]=mm[irow,icol] >> } >> } >> mm mm >>> [,1] [,2] [,3] [,4] [,5] >>> [1,]12345 >>> [2,]6789 10 >>> [3,] 11 12 13 14 15 >>> >> xmm xmm >>> [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] >> [,13] [,14] [,15] >>> [1,]111222333 4 4 4 >> 5 5 5 >>> [2,]111222333 4 4 4 >> 5 5 5 >>> [3,]111222333 4 4 4 >> 5 5 5 >>> [4,]666777888 9 9 9 >> 101010 >>> [5,]666777888 9 9 9 >> 101010 >>> [6,]666777888 9 9 9 >> 101010 >>> [7,] 11 11 11 12 12 12 13 13 13141414 >> 151515 >>> [8,] 11 11 11 12 12 12 13 13 13141414 >> 151515 >>> [9,] 11 11 11 12 12 12 13 13 13141414 >> 151515 >> >> I tried various rep with matrix, but don't get the right result. >>
Re: [R] expand gridded matrix to higher resolution
You probably ought to be using the raster package. See the CRAN Spatial Task View. -- Sent from my phone. Please excuse my brevity. On July 5, 2017 12:20:28 AM PDT, "Anthoni, Peter (IMK)"wrote: >Hi all, >(if me email goes out as html, than my email client don't do as told, >and I apologies already.) > >We need to downscale climate data and therefore first need to expand >the climate from 0.5deg to the higher resolution 10min, before we can >add high resolution deviations. We basically need to have the original >data at each gridcell replicated into 3x3 gridcells. >A simple for loop can do this, but I could need a faster procedure. >Anybody know a faster way? Is there package than can do what we need >already? >I tried matrix with rep, but I am missing some magic there, since it >doesn't do what we need. >replicate might be promising, but then still need to rearrange the >output into the column and row format we need. > >A simple example: >mm=matrix(1:15,nrow=3,byrow = T) >xmm=matrix(NA,nrow=nrow(mm)*3,ncol=ncol(mm)*3) >for(icol in 1:ncol(mm)) { > for(irow in 1:nrow(mm)) { >xicol=(icol-1)*3 +c(1:3) >xirow=(irow-1)*3 +c(1:3) >xmm[xirow,xicol]=mm[irow,icol] > } >} >mm >> > mm >> [,1] [,2] [,3] [,4] [,5] >> [1,]12345 >> [2,]6789 10 >> [3,] 11 12 13 14 15 >> >xmm >> > xmm >> [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] >[,13] [,14] [,15] >> [1,]111222333 4 4 4 > 5 5 5 >> [2,]111222333 4 4 4 > 5 5 5 >> [3,]111222333 4 4 4 > 5 5 5 >> [4,]666777888 9 9 9 > 101010 >> [5,]666777888 9 9 9 > 101010 >> [6,]666777888 9 9 9 > 101010 >> [7,] 11 11 11 12 12 12 13 13 13141414 > 151515 >> [8,] 11 11 11 12 12 12 13 13 13141414 > 151515 >> [9,] 11 11 11 12 12 12 13 13 13141414 > 151515 > >I tried various rep with matrix, but don't get the right result. >xmm2=matrix(rep(rep(mm,each=3),times=3),nrow=nrow(mm)*3,ncol=ncol(mm)*3,byrow >= F) >> identical(xmm,xmm2) >[1] FALSE > >rr=replicate(3,rep(t(mm),each=3)) >rr >> > rr >> [,1] [,2] [,3] >> [1,]111 >> [2,]111 >> [3,]111 >> [4,]222 >> [5,]222 >> [6,]222 >> [7,]333 >> ... >identical(xmm,matrix(rr,ncol=15,nrow=9,byrow=T)) >> > identical(xmm,matrix(rr,ncol=15,nrow=9,byrow=T)) >> [1] FALSE > >Many thanks for any advice. > >cheers >Peter > >__ >R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see >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 -- To UNSUBSCRIBE and more, see 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] expand gridded matrix to higher resolution
Hi Jim, thanks that works like a charm. cheers Peter > On 5. Jul 2017, at 12:01, Jim Lemonwrote: > > Hi Peter, > > apply(t(apply(mm,1,rep,each=3)),2,rep,each=3) > > Jim > > On Wed, Jul 5, 2017 at 5:20 PM, Anthoni, Peter (IMK) > wrote: >> Hi all, >> (if me email goes out as html, than my email client don't do as told, and I >> apologies already.) >> >> We need to downscale climate data and therefore first need to expand the >> climate from 0.5deg to the higher resolution 10min, before we can add high >> resolution deviations. We basically need to have the original data at each >> gridcell replicated into 3x3 gridcells. >> A simple for loop can do this, but I could need a faster procedure. Anybody >> know a faster way? Is there package than can do what we need already? >> I tried matrix with rep, but I am missing some magic there, since it doesn't >> do what we need. >> replicate might be promising, but then still need to rearrange the output >> into the column and row format we need. >> >> A simple example: >> mm=matrix(1:15,nrow=3,byrow = T) >> xmm=matrix(NA,nrow=nrow(mm)*3,ncol=ncol(mm)*3) >> for(icol in 1:ncol(mm)) { >> for(irow in 1:nrow(mm)) { >>xicol=(icol-1)*3 +c(1:3) >>xirow=(irow-1)*3 +c(1:3) >>xmm[xirow,xicol]=mm[irow,icol] >> } >> } >> mm mm >>> [,1] [,2] [,3] [,4] [,5] >>> [1,]12345 >>> [2,]6789 10 >>> [3,] 11 12 13 14 15 >>> >> xmm xmm >>> [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13] >>> [,14] [,15] >>> [1,]111222333 4 4 4 5 >>> 5 5 >>> [2,]111222333 4 4 4 5 >>> 5 5 >>> [3,]111222333 4 4 4 5 >>> 5 5 >>> [4,]666777888 9 9 910 >>> 1010 >>> [5,]666777888 9 9 910 >>> 1010 >>> [6,]666777888 9 9 910 >>> 1010 >>> [7,] 11 11 11 12 12 12 13 13 1314141415 >>> 1515 >>> [8,] 11 11 11 12 12 12 13 13 1314141415 >>> 1515 >>> [9,] 11 11 11 12 12 12 13 13 1314141415 >>> 1515 >> >> I tried various rep with matrix, but don't get the right result. >> xmm2=matrix(rep(rep(mm,each=3),times=3),nrow=nrow(mm)*3,ncol=ncol(mm)*3,byrow >> = F) >>> identical(xmm,xmm2) >> [1] FALSE >> >> rr=replicate(3,rep(t(mm),each=3)) >> rr rr >>> [,1] [,2] [,3] >>> [1,]111 >>> [2,]111 >>> [3,]111 >>> [4,]222 >>> [5,]222 >>> [6,]222 >>> [7,]333 >>> ... >> identical(xmm,matrix(rr,ncol=15,nrow=9,byrow=T)) identical(xmm,matrix(rr,ncol=15,nrow=9,byrow=T)) >>> [1] FALSE >> >> Many thanks for any advice. >> >> cheers >> Peter >> >> __ >> R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see >> 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 -- To UNSUBSCRIBE and more, see 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] expand gridded matrix to higher resolution
Hi Peter, apply(t(apply(mm,1,rep,each=3)),2,rep,each=3) Jim On Wed, Jul 5, 2017 at 5:20 PM, Anthoni, Peter (IMK)wrote: > Hi all, > (if me email goes out as html, than my email client don't do as told, and I > apologies already.) > > We need to downscale climate data and therefore first need to expand the > climate from 0.5deg to the higher resolution 10min, before we can add high > resolution deviations. We basically need to have the original data at each > gridcell replicated into 3x3 gridcells. > A simple for loop can do this, but I could need a faster procedure. Anybody > know a faster way? Is there package than can do what we need already? > I tried matrix with rep, but I am missing some magic there, since it doesn't > do what we need. > replicate might be promising, but then still need to rearrange the output > into the column and row format we need. > > A simple example: > mm=matrix(1:15,nrow=3,byrow = T) > xmm=matrix(NA,nrow=nrow(mm)*3,ncol=ncol(mm)*3) > for(icol in 1:ncol(mm)) { > for(irow in 1:nrow(mm)) { > xicol=(icol-1)*3 +c(1:3) > xirow=(irow-1)*3 +c(1:3) > xmm[xirow,xicol]=mm[irow,icol] > } > } > mm >> > mm >> [,1] [,2] [,3] [,4] [,5] >> [1,]12345 >> [2,]6789 10 >> [3,] 11 12 13 14 15 >> > xmm >> > xmm >> [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13] >> [,14] [,15] >> [1,]111222333 4 4 4 5 >> 5 5 >> [2,]111222333 4 4 4 5 >> 5 5 >> [3,]111222333 4 4 4 5 >> 5 5 >> [4,]666777888 9 9 910 >> 1010 >> [5,]666777888 9 9 910 >> 1010 >> [6,]666777888 9 9 910 >> 1010 >> [7,] 11 11 11 12 12 12 13 13 1314141415 >> 1515 >> [8,] 11 11 11 12 12 12 13 13 1314141415 >> 1515 >> [9,] 11 11 11 12 12 12 13 13 1314141415 >> 1515 > > I tried various rep with matrix, but don't get the right result. > xmm2=matrix(rep(rep(mm,each=3),times=3),nrow=nrow(mm)*3,ncol=ncol(mm)*3,byrow > = F) >> identical(xmm,xmm2) > [1] FALSE > > rr=replicate(3,rep(t(mm),each=3)) > rr >> > rr >> [,1] [,2] [,3] >> [1,]111 >> [2,]111 >> [3,]111 >> [4,]222 >> [5,]222 >> [6,]222 >> [7,]333 >> ... > identical(xmm,matrix(rr,ncol=15,nrow=9,byrow=T)) >> > identical(xmm,matrix(rr,ncol=15,nrow=9,byrow=T)) >> [1] FALSE > > Many thanks for any advice. > > cheers > Peter > > __ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > 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 -- To UNSUBSCRIBE and more, see 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] expand gridded matrix to higher resolution
Hi all, (if me email goes out as html, than my email client don't do as told, and I apologies already.) We need to downscale climate data and therefore first need to expand the climate from 0.5deg to the higher resolution 10min, before we can add high resolution deviations. We basically need to have the original data at each gridcell replicated into 3x3 gridcells. A simple for loop can do this, but I could need a faster procedure. Anybody know a faster way? Is there package than can do what we need already? I tried matrix with rep, but I am missing some magic there, since it doesn't do what we need. replicate might be promising, but then still need to rearrange the output into the column and row format we need. A simple example: mm=matrix(1:15,nrow=3,byrow = T) xmm=matrix(NA,nrow=nrow(mm)*3,ncol=ncol(mm)*3) for(icol in 1:ncol(mm)) { for(irow in 1:nrow(mm)) { xicol=(icol-1)*3 +c(1:3) xirow=(irow-1)*3 +c(1:3) xmm[xirow,xicol]=mm[irow,icol] } } mm > > mm > [,1] [,2] [,3] [,4] [,5] > [1,]12345 > [2,]6789 10 > [3,] 11 12 13 14 15 > xmm > > xmm > [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13] > [,14] [,15] > [1,]111222333 4 4 4 5 > 5 5 > [2,]111222333 4 4 4 5 > 5 5 > [3,]111222333 4 4 4 5 > 5 5 > [4,]666777888 9 9 910 > 1010 > [5,]666777888 9 9 910 > 1010 > [6,]666777888 9 9 910 > 1010 > [7,] 11 11 11 12 12 12 13 13 1314141415 > 1515 > [8,] 11 11 11 12 12 12 13 13 1314141415 > 1515 > [9,] 11 11 11 12 12 12 13 13 1314141415 > 1515 I tried various rep with matrix, but don't get the right result. xmm2=matrix(rep(rep(mm,each=3),times=3),nrow=nrow(mm)*3,ncol=ncol(mm)*3,byrow = F) > identical(xmm,xmm2) [1] FALSE rr=replicate(3,rep(t(mm),each=3)) rr > > rr > [,1] [,2] [,3] > [1,]111 > [2,]111 > [3,]111 > [4,]222 > [5,]222 > [6,]222 > [7,]333 > ... identical(xmm,matrix(rr,ncol=15,nrow=9,byrow=T)) > > identical(xmm,matrix(rr,ncol=15,nrow=9,byrow=T)) > [1] FALSE Many thanks for any advice. cheers Peter __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.