How to find out if a data frame has automatic row names?

Hi all,

I just stumbled across the difference between explicit and 'automatic' row
names for data frames. Is there a quick way to find out if a data frame has
automatic row names or not? Of course I know this for data frames that I
create myself, but what if a function needs to extract this information for
an argument?

The following example shows that this property is not easily visible:

# data frame with automatic row names
df <- data.frame(a=1:2, b=3:4)
# data frame with explicit row names
df2 <- data.frame(a=1:2, b=3:4, row.names=1:2)
# printing does not reveal any difference
df

?a b
1 1 3
2 2 4

df2

?a b
1 1 3
2 2 4

# both are considered equal
all.equal(df, df2)

[1] TRUE

identical(df, df2)

[1] TRUE

# calling rownames gives the same result
rownames(df)

[1] "1" "2"

rownames(df2)

[1] "1" "2"

# when converting to a matrix, it makes a difference
as.matrix(df)

? ?a b
[1,] 1 3
[2,] 2 4

as.matrix(df2)

?a b
1 1 3
2 2 4

Thanks for any suggestion,

Andreas

--
Andreas Borg
Medizinische Informatik

Hi all,

I just stumbled across the difference between explicit and 'automatic'
row names for data frames. Is there a quick way to find out if a data
frame has automatic row names or not? Of course I know this for data
frames that I create myself, but what if a function needs to extract
this information for an argument?

The following example shows that this property is not easily visible:

  >  # data frame with automatic row names
  >  df<- data.frame(a=1:2, b=3:4)
  >  # data frame with explicit row names
  >  df2<- data.frame(a=1:2, b=3:4, row.names=1:2)
  >  # printing does not reveal any difference
  >  df

   a b
1 1 3
2 2 4

  >  df2

   a b
1 1 3
2 2 4

  >  # both are considered equal
  >  all.equal(df, df2)

[1] TRUE

  >  identical(df, df2)

[1] TRUE

  >  # calling rownames gives the same result
  >  rownames(df)

[1] "1" "2"

  >  rownames(df2)

[1] "1" "2"

  >  # when converting to a matrix, it makes a difference
  >  as.matrix(df)

      a b
[1,] 1 3
[2,] 2 4

  >  as.matrix(df2)

   a b
1 1 3
2 2 4

Thanks for any suggestion,

Andreas

Check ?.row_names_info

Peter Ehlers

Andreas