Btukfyl

I have a whole bunch of matrices that I want to assign column and row names to based on a column of names that are held in another column (Names) in a dataframe (months) I have.

Each matrix I want to assign names to corresponds to a concatenated "partner" string and metric held here e.g. one of the matrices is called "facebook_Total_Completed_Rentals":

Metrics = c('_Total_Completed_Rentals','_Registered_Completed_rentals',

'_Registered_renters','_New_Registrations')



Partners = c('facebook', 'altrooz', 'adperio', 'snapchat', 'apple', 'google')    

I want to loop through these two to assign the names to the matrices in this way to avoid having to manually code out all the colnames and row.names statements:

for(i in 1:length(Metrics)){

  for(j in 1:length(Partners)){

  temp = paste0(Partners[j],Metrics[i])

  colnames(temp)= months$Names[1:(months_cut-1)]

  row.names(temp) = months$Names[1:(months_cut-1)]

  }

}  

But I get the error:

Error in `colnames<-`(`*tmp*`, value = months$Names[1:(months_cut - 1)]) : 

attempt to set 'colnames' on an object with less than two dimensions

Please help

Consider the following data.frames :

iris_a  <- iris[1:2,]

iris_b  <- iris[3:4,]

iris_c  <- iris[5:6,]

The clean way to do what you want is to put them in a list and apply changes on it :

rnms <- c("foo","bar")

your_list <- lapply(list(iris_a = iris_a,iris_b = iris_b, iris_c = iris_c), function(x){

  rownames(x) <- rnms

  x

})

Then if necessary you can extract its elements back to your local environment :

list2env(your_list, environment())

# Sepal.Length Sepal.Width Petal.Length Petal.Width Species

# foo          5.1         3.5          1.4         0.2  setosa

# bar          4.9         3.0          1.4         0.2  setosa

A bit more compact using tidyverse :

your_list <- map(lst(iris_a, iris_b, iris_c), `rownames<-`, rnms)

While @Moody's answer is probably the cleaner option, I thought I'd provide code that does what you originally wanted it to do:

Let's create 4 matrices whose names are combinations of 2 elements:

a_1 <- a_2 <- b_1 <- b_2 <- matrix(1:20, ncol = 5)

Vector of row names and column names:

rnames <- paste0("row_", 1:4)

cnames <- paste0("col_", 1:5)

Get the name elements:

x <- c("a", "b")

y <- 1:2

For your loop, you need to realise that temp is just a character string so colnames(temp) will not work. You want to create a string that includes the entire command, parse it as an expression, and then evaluate it:

for (i in x) {

  for (j in y) {

    temp <- paste(i, j, sep = "_")

    eval(parse(text = paste0("dimnames(", temp, ") <- list(rnames, cnames)")))

  }

}

Note. dimnames(temp) is quicker than rownames and colnames.