cbind warnings : row names were found from a short variable and have been discarded

Question

I have below line of code for cbind, but I am getting a warning message everytime. Though the code still functions as it should be, is there any way to resolve the warning?

dateset = subset(all_data[,c("VAR1","VAR2","VAR3","VAR4","VAR5","RATE1","RATE2","RATE3")])
dateset = cbind(dateset[c(1,2,3,4,5)],stack(dateset[,-c(1,2,3,4,5)]))

Warnings :

Warning message:
   In data.frame(..., check.names = FALSE) :
        row names were found from a short variable and have been discarded

Thanks in advance!

Answer 1

I'm guessing your data.frame has row.names:

A <- data.frame(a = c("A", "B", "C"), 
                b = c(1, 2, 3), 
                c = c(4, 5, 6), 
                row.names=c("A", "B", "C"))

cbind(A[1], stack(A[-1]))
#   a values ind
# 1 A      1   b
# 2 B      2   b
# 3 C      3   b
# 4 A      4   c
# 5 B      5   c
# 6 C      6   c
# Warning message:
# In data.frame(..., check.names = FALSE) :
#   row names were found from a short variable and have been discarded

---

What's happening here is that since you can't by default have duplicated row.names in a data.frame and since you don't tell R at any point to duplicate the row.names when recycling the first column to the same number of rows of the stacked column, R just discards the row.names.

Compare with a similar data.frame, but one without row.names:

B <- data.frame(a = c("A", "B", "C"), 
                b = c(1, 2, 3), 
                c = c(4, 5, 6))

cbind(B[1], stack(B[-1]))
#   a values ind
# 1 A      1   b
# 2 B      2   b
# 3 C      3   b
# 4 A      4   c
# 5 B      5   c
# 6 C      6   c

---

Alternatively, you can set row.names = NULL in your cbind statement:

cbind(A[1], stack(A[-1]), row.names = NULL)
#   a values ind
# 1 A      1   b
# 2 B      2   b
# 3 C      3   b
# 4 A      4   c
# 5 B      5   c
# 6 C      6   c

---

If your original row.names are important, you can also add them back in with:

cbind(rn = rownames(A), A[1], stack(A[-1]), row.names = NULL)
#   rn a values ind
# 1  A A      1   b
# 2  B B      2   b
# 3  C C      3   b
# 4  A A      4   c
# 5  B B      5   c
# 6  C C      6   c