CODEWITHSUNDEEP
r
Jessica
Jessica

Reputation: 461

How to solve dimnames error for matrix in R

I am using this code to make a matrix:

input1 <- read_excel("input.xlsx")
MAT <- as.matrix(input1[,2:400])
Height <- MAT[1,]
Expres <- MAT[2:36000,]
row.names(Expres) <- input1[2:36000, 1]

My input data has 400 samples with 36000 genes and Height in row 2 and the data subset looks like:

Gene Sample1 Sample2 -- Sample400
Height 5 10 -- 7 
Gene1 0.5 0.6 -- 0.9
Gene2 0.4 0.7 -- 0.3
Gene36000 0.5 0.6 -- 0.1

but I am getting an error for the last line:

Error in dimnames(x) <- dn : 
  length of 'dimnames' [1] not equal to array extent

I will appreciate any help. Thank you!

Upvotes: 0

Views: 665

Answers (1)

r2evans
r2evans

Reputation: 160792

Assuming you are using readxl::read_excel, it's returning a tibble, and [ indexing is not doing what you expect.

Your expectation works with a data.frame:

mtcars[2:3,1]
# [1] 21.0 22.8

but not with a tibble:

as_tibble(mtcars)[2:3,1]
# # A tibble: 2 x 1
#     mpg
#   <dbl>
# 1  21  
# 2  22.8

as_tibble(mtcars)[[1]][2:3]
# [1] 21.0 22.8

Instead, use

row.names(Expres) <- input1[[1]][2:36000]

Incidentally, the expected behavior can be forced in both directions

mtcars[2:3, 1, drop = FALSE]
#                mpg
# Mazda RX4 Wag 21.0
# Datsun 710    22.8
as_tibble(mtcars)[2:3, 1, drop = TRUE]
# [1] 21.0 22.8

Sadly, the use of drop= works with data.frame and tbl_df, but not with the third data.frame variant data.table:

library(data.table)
as.data.table(mtcars)[2:3, 1, drop = TRUE]
#     mpg
# 1: 21.0
# 2: 22.8
as.data.table(mtcars)[2:3, 1, drop = FALSE]
#     mpg
# 1: 21.0
# 2: 22.8
as.data.table(mtcars)[[1]][2:3]
# [1] 21.0 22.8

(A source of many hours of trouble-shooting for me some time ago.)

Upvotes: 2

Related Questions