CODEWITHSUNDEEP
r
Jack
Jack

Reputation: 722

return type of order() in R

I'm a newbie in R and I'm trying to debug a special case in my code. In the following lines what I'm doing is getting a vector x sorted, based a column that surely exists. Naturally, I would expect sorted.data to be the same and in most cases it is. 

print (paste("x=",typeof(x)))
print (x)
sorted.data <- x[order(x[[max.column]]), ]
print (paste("type of sorted=", typeof(sorted.data)))
print (sorted.data)

In most cases the output of the above code is:

[1] "x= list"
    V1  V5
8   10   0
16   7   0
18   7   0
20   1  96
24   9   0

[1] "type of sorted= list"
    V1  V5
8   10   0
16   7   0
18   7   0
24   9   0
31   5   0

But at some point, the output is:

[1] "x= list"    # Expected. a vector with 1 column
    V1
8   10
16   7
18   7
24   9
26  10

[1] "type of sorted= integer"  # This shouldn't happen
 [1]  0  0  0  1  1  1  1  2  2  3  3  4  4  4

What could be going wrong? Why the return value changes?

Upvotes: 1

Views: 77

Answers (1)

r2evans
r2evans

Reputation: 160607

First, your data looks very much like a data.frame, which is very much like a list. If it were truly not a data.frame, then your comma-indexing (x[order(x[[max.column]]), ]) should not work, so I'm assuming it's a data.frame.

One thing I find occasionally frustrating with data.frame indexing (and arrays, too), the default behavior is that "the result is coerced to the lowest possible dimension" (extracted from help("[")). This means that when the indexing results in only one column (or row), it is returned as a vector. To preempt this, add ,drop=FALSE to your indexing:

x[order(x[[max.column]]),,drop=FALSE]

(Thanks to Rui Barradas for correcting my errant comment above.)

Upvotes: 2

Related Questions