CODEWITHSUNDEEP
rfor-loopdataframematrix
Seymoo
Seymoo

Reputation: 189

Add columns to a data frame by for loop

I have a matrix data and a list. I would like to find matches between each vector in my list with row names of my matrix, calculate their mean and add those values in data frame df with new coumn of data frame having the same name as the names of each vector in the list. I have done it manually, but I am wondering how can I do it in a for loop manner with more efficient codes.

data <- matrix(runif(75, 5.0, 10), nrow = 15, ncol = 5)
rownames(data) <- paste0("GENE",1:15)

colnames(data) <- paste0("COL",1:5)
list <- list(n = c("GENE1","GENE2","GENE3"), s = c("GENE4", "GENE5","GENE6","GENE7"),b = c("GENE8","GENE10", "GENE12", "GENE15"))

df <- data.frame(row.names = colnames(data))
df$n <- matrixStats::rowMeans2(t(data[intersect(row.names(data),list$n),]))
df$s <- matrixStats::rowMeans2(t(data[intersect(row.names(data),list$s),]))
df$b <- matrixStats::rowMeans2(t(data[intersect(row.names(data),list$b),]))

Upvotes: 0

Views: 75

Answers (1)

Ronak Shah
Ronak Shah

Reputation: 388817

If you want to optimise instead of loops try lapply in base R, where we subset rows based on rownames in each element of list and then take mean of every column using colMeans. 

t(do.call("rbind", lapply(lst, function(x) 
                   colMeans(data[rownames(data) %in% x,]))))


#            n        s        b
#COL1 7.242129 7.667626 6.980115
#COL2 7.317233 6.297818 6.186642
#COL3 6.709917 7.061652 7.552923
#COL4 7.773472 6.741069 7.765780
#COL5 7.039789 6.584206 7.569894

data

set.seed(1234)
data <- matrix(runif(75, 5.0, 10), nrow = 15, ncol = 5)
rownames(data) <- paste0("GENE",1:15)

colnames(data) <- paste0("COL",1:5)
lst <- list(n=c("GENE1","GENE2","GENE3"), s = c("GENE4", "GENE5","GENE6","GENE7"),
    b = c("GENE8","GENE10", "GENE12", "GENE15"))

Upvotes: 1

Related Questions