Cade, Brian
2014-Apr-15 20:26 UTC
[R] indexing names for looping across computations done on pairs of matrices
So I know I must be missing something simple and obvious for the following
data manipulation where I have (in this example) 11 pairs of matrices
(gs4.0 to gs4.100 and ps1.0 to ps1.100) from some population simulations
(all with same dimensions) where I want to get some summary statistics on
the products of the cells in a pair (e.g., gs4.0 * ps1.0). The code I
wrote below works fine, but it seems like there ought to be a simple way to
index the extensions on the names (.0 to .100) in a for loop to simplify
this code greatly. I've spent some time trying various things using
paste() and assign() and have had no success.
mean.comb <- as.matrix(0:10,nrow = 11, ncol=1)
mean.comb <- cbind(mean.comb,0)
###to see list of files created
gs4files <- ls(pattern="gs4.*0")
ps1files <- ls(pattern="ps1.*0")
mean.comb[1,2] <- mean(apply(gs4.0 * ps1.0,1,sum))
mean.comb[2,2] <- mean(apply(gs4.10 * ps1.10,1,sum))
mean.comb[3,2] <- mean(apply(gs4.20 * ps1.20,1,sum))
mean.comb[4,2] <- mean(apply(gs4.30 * ps1.30,1,sum))
mean.comb[5,2] <- mean(apply(gs4.40 * ps1.40,1,sum))
mean.comb[6,2] <- mean(apply(gs4.50 * ps1.50,1,sum))
mean.comb[7,2] <- mean(apply(gs4.60 * ps1.60,1,sum))
mean.comb[8,2] <- mean(apply(gs4.70 * ps1.70,1,sum))
mean.comb[9,2] <- mean(apply(gs4.80 * ps1.80,1,sum))
mean.comb[10,2] <- mean(apply(gs4.90 * ps1.90,1,sum))
mean.comb[11,2] <- mean(apply(gs4.100 * ps1.100,1,sum))
mean.comb<- data.frame(mean.comb)
colnames(mean.comb) <- c("simulation",
"mean.horses.removed")
Brian
Brian S. Cade, PhD
U. S. Geological Survey
Fort Collins Science Center
2150 Centre Ave., Bldg. C
Fort Collins, CO 80526-8818
email: cadeb@usgs.gov <brian_cade@usgs.gov>
tel: 970 226-9326
[[alternative HTML version deleted]]
arun
2014-Apr-16 02:11 UTC
[R] indexing names for looping across computations done on pairs of matrices
Hi,
If the two pairs of matrices are in a list:
set.seed(42)
lst1 <- lapply(1:11, function(x) matrix(sample(40, 20, replace=TRUE), 5,4))
names(lst1) <- paste0("gs",
paste0("4.",seq(0,100,by=10)))
set.seed(585)
lst2 <- lapply(1:11, function(x) matrix(sample(40, 20, replace=TRUE), 5,4))
names(lst2) <- paste0("ps",
paste0("1.",seq(0,100,by=10)))
mean.comb <- data.frame(simulation=0:10,
mean.horses.removed=sapply(seq_along(lst1),function(i)
mean(rowSums(lst1[[i]]*lst2[[i]]))))?
#or
# if the datasets are standalone with no particular order
#just for demonstration
attach(lst1)
attach(lst2)
vec1 <- sample(paste0("gs",
paste0("4.",seq(0,100,by=10))), 11, replace=FALSE)
vec2 <- sample(paste0("ps",
paste0("1.",seq(0,100,by=10))), 11, replace=FALSE)
vec1New <- vec1[order(as.numeric(gsub(".*\\.","",vec1)))]
vec2New <- vec2[order(as.numeric(gsub(".*\\.","",vec2)))]
mean.comb2 <- data.frame(simulation=0:10,
mean.horses.removed=sapply(seq_along(vec1New),function(i)
mean(rowSums(get(vec1New[i])* get(vec2New[i])))))
identical(mean.comb,mean.comb2)
#[1] TRUE
A.K.
On Tuesday, April 15, 2014 4:30 PM, "Cade, Brian" <cadeb at
usgs.gov> wrote:
So I know I must be missing something simple and obvious for the following
data manipulation where I have (in this example) 11 pairs of matrices
(gs4.0 to gs4.100 and ps1.0 to ps1.100) from some population simulations
(all with same dimensions) where I want to get some summary statistics on
the products of the cells in a pair (e.g., gs4.0 * ps1.0).? The code I
wrote below works fine, but it seems like there ought to be a simple way to
index the extensions on the names (.0 to .100) in a for loop to simplify
this code greatly.? I've spent some time trying various things using
paste() and assign() and have had no success.
mean.comb <- as.matrix(0:10,nrow = 11, ncol=1)
mean.comb <- cbind(mean.comb,0)
###to see list of files created
gs4files <- ls(pattern="gs4.*0")
ps1files <- ls(pattern="ps1.*0")
mean.comb[1,2] <- mean(apply(gs4.0 * ps1.0,1,sum))
mean.comb[2,2] <- mean(apply(gs4.10 * ps1.10,1,sum))
mean.comb[3,2] <- mean(apply(gs4.20 * ps1.20,1,sum))
mean.comb[4,2] <- mean(apply(gs4.30 * ps1.30,1,sum))
mean.comb[5,2] <- mean(apply(gs4.40 * ps1.40,1,sum))
mean.comb[6,2] <- mean(apply(gs4.50 * ps1.50,1,sum))
mean.comb[7,2] <- mean(apply(gs4.60 * ps1.60,1,sum))
mean.comb[8,2] <- mean(apply(gs4.70 * ps1.70,1,sum))
mean.comb[9,2] <- mean(apply(gs4.80 * ps1.80,1,sum))
mean.comb[10,2] <- mean(apply(gs4.90 * ps1.90,1,sum))
mean.comb[11,2] <- mean(apply(gs4.100 * ps1.100,1,sum))
mean.comb<- data.frame(mean.comb)
colnames(mean.comb) <- c("simulation",
"mean.horses.removed")
Brian
Brian S. Cade, PhD
U. S. Geological Survey
Fort Collins Science Center
2150 Centre Ave., Bldg. C
Fort Collins, CO? 80526-8818
email:? cadeb at usgs.gov <brian_cade at usgs.gov>
tel:? 970 226-9326
??? [[alternative HTML version deleted]]
______________________________________________
R-help at 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.