Unable to set column names to a subset of a dataframe

Question

I run the following code, p is the dataframe loaded.

a <- sort(table(p$Title))
a1 <- as.data.frame(a)
tail(a1, 7)

                     a
Maths               732
Science             737
Physics             737
Chemistry           776
Social Science      905
null              57374
                  88117

I want to do some manipulations on the above dataframe result. I want to add column names to the dataframe. I tried the colnames function.

colnames(a1) <- c("category", "count")

I get the below error:

Error in `colnames<-`(`*tmp*`, value = c("category", "count")) : 
    attempt to set 'colnames' on an object with less than two dimensions

Please suggest.

As @Jaap said, it seems you have row names there. Try `a1$category <- row.names(a1)` maybe (unless it was a typo as @Richard suggests) — David Arenburg, Aug 02 '15 at 05:42
@RichardScriven Here is what I tried. colnames(a1)<-data.frame("category", "count") Error in `colnames<-`(`*tmp*`, value = list(X.category. = 1L, X.count. = 1L)) : 'names' attribute [2] must be the same length as the vector [1] — RpB, Aug 02 '15 at 05:47
Yeah, you need to shift the row names into a column, as mentioned above. — Rich Scriven, Aug 02 '15 at 05:50

Jaap · Accepted Answer · 2015-08-08T19:18:03.450

As I said in the comments to your question, the categories are rownames. A reproducible example:

# create dataframe p
x <- c("Maths","Science","Physics","Chemistry","Social Science","Languages","Economics","History")
set.seed(1)
p <- data.frame(title=sample(x, 100, replace=TRUE), y="some arbitrary value")

# create the data.frame as you did
a <- sort(table(p$title))
a1 <- as.data.frame(a)

The resulting dataframe:

> a1
                a
Social Science  6
Maths           9
History        10
Science        11
Physics        12
Languages      15
Economics      17
Chemistry      20

Looking at the dimensions of dataframe a1, you get this:

> dim(a1)
[1] 8 1

which means that your dataframe has 8 rows and 1 column. Trying to assign two columnnames to the a1 dataframe will hence result in an error.

You can solve your problem in two ways:

1: assign just 1 columnname with colnames(a1) <- c("count")

2: convert the rownames to a category column and then assign the columnnames:

a1$category <- row.names(a1)
colnames(a1) <- c("count","category")

The resulting dataframe:

> a1
               count       category
Social Science     6 Social Science
Maths              9          Maths
History           10        History
Science           11        Science
Physics           12        Physics
Languages         15      Languages
Economics         17      Economics
Chemistry         20      Chemistry

You can remove the rownames with rownames(a1) <- NULL. This gives:

> a1

  count       category
1     6 Social Science
2     9          Maths
3    10        History
4    11        Science
5    12        Physics
6    15      Languages
7    17      Economics
8    20      Chemistry

Unable to set column names to a subset of a dataframe

1 Answers1