R：在 for 循环中命名矩阵行/列在控制台中有效，但在编织时无效

Question

I am attempting to use a for loop to rename the rows and columns of a matrix via the following code:

Z = matrix(c(1:120), 15, 8)
for (i in 1:15) {
  rownames(Z)[i] = paste0("[", format((-1)*(4-(i/10)), nsmall=1), "]")
}
for (j in 1:8) {
  colnames(Z)[j] = paste0("[", (j-1)/100, "]")
}

This works as expected when entered into the console, and produces the following matrix Z :

       [.00] [.01] [.02] [.03] [.04] [.05] [.06] [.07]
[-3.9]     1    16    31    46    61    76    91   106
[-3.8]     2    17    32    47    62    77    92   107
[-3.7]     3    18    33    48    63    78    93   108
[-3.6]     4    19    34    49    64    79    94   109
[-3.5]     5    20    35    50    65    80    95   110
[-3.4]     6    21    36    51    66    81    96   111
[-3.3]     7    22    37    52    67    82    97   112
[-3.2]     8    23    38    53    68    83    98   113
[-3.1]     9    24    39    54    69    84    99   114
[-3.0]    10    25    40    55    70    85   100   115
[-2.9]    11    26    41    56    71    86   101   116
[-2.8]    12    27    42    57    72    87   102   117
[-2.7]    13    28    43    58    73    88   103   118
[-2.6]    14    29    44    59    74    89   104   119
[-2.5]    15    30    45    60    75    90   105   120

However, I need to include this in an R Markdown document that knits to PDF, as in this MWE:

---
title: "Title"
author: "Author"
date: "1970-01-01"
output: pdf_document
---

```{r}
Z = matrix(c(1:120), 15, 8)
for (i in 1:15) {
  rownames(Z)[i] = paste0("[", format((-1)*(4-(i/10)), nsmall=1), "]")
}
for (j in 1:8) {
  colnames(Z)[j] = paste0("[", (j-1)/100, "]")
}
```

When trying to knit this document, R Markdown returns the error

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

and the execution halts.

I found that the above error has been extensively discussed, but that the cause of the error is when attempting to supply too few or too many names for the number of rows or columns. However, that should not be the case here, and since the code does not return this error when entered into the console, I have been unable to find a cause for Markdown rejecting it.

(I realize that assigning vectors to the row and column names is preferred, especially since the names are known; but this is for a university class assignment which is intended to familiarize me with loops, and so the assignment requires that everything must be done inside loops.)

I am using R version 3.6.2 and RStudio version 1.2.5033 on Windows 10 version 1909 build 18363.1082. Any help is appreciated!

Edit 1: Output of sessionInfo() as requested by @Wil:

R version 3.6.2 (2019-12-12)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 18363)

Matrix products: default

Random number generation:
 RNG:     Mersenne-Twister 
 Normal:  Inversion 
 Sample:  Rounding 
 
locale:
[1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252   
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C                          
[5] LC_TIME=English_United States.1252    

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

loaded via a namespace (and not attached):
 [1] compiler_3.6.2  htmltools_0.4.0 tools_3.6.2     yaml_2.2.0      Rcpp_1.0.3     
 [6] rmarkdown_2.0   knitr_1.26      digest_0.6.23   xfun_0.12       rlang_0.4.4    
[11] evaluate_0.14  

Answer 1

For a matrix Z without any defined row or column names, the value of rownames(Z) and colnames(Z) is NULL . You're getting this error because you're trying to subset [i] into NULL , which doesn't work.

Z = matrix(c(1:120), 15, 8)
rownames(Z) # NULL
rownames(Z)[1] # NULL 
rownames(Z)[1] <- "foo" 
# Error in dimnames(x) <- dn : length of 'dimnames' [1] not equal to array extent

Note that this is discussed in the base R documentation for colnames() / rownames() :

constructions such as

rownames(x)[3] <- "c"

may not work unless x already has dimnames, since this will create a length-3 value from the NULL value of rownames(x).

You can assign rownames with a vectorized input, such as:

rownames(Z) <- paste0("[", format((-1)*(4-(1:15/10)), nsmall=1), "]")

But from your post it sounds like this exercise is focused on for -loops. Per the docs, this isn't a viable option for setting names on a new matrix.

If you can't use a purely vectorized approach, the closest solution involving a for -loop that I can think of is to loop the rowname values into a vector, then assign the vector as rownames(Z) . But this seems like a needless middle step, so I'm not sure it's really a useful pattern to learn.

new_rownames <- vector(mode = "character", length = nrow(Z))

for (i in 1:15) {
  new_rownames[i] = paste0("[", format((-1)*(4-(i/10)), nsmall=1), "]")
}

rownames(Z) <- new_rownames

Another option might be to assign an empty set of values to rownames/colnames for Z :

Z = matrix(c(1:120), 15, 8, dimnames = list(1:15, 1:8))

Then you could use your original loops as intended.