rbind: inconsistent behaviour with empty data frames?

What version of R are you using?

What version of R are you using?
I cannot replicate your results with R 2.15.2:

df1 <- data.frame()
df2 <- data.frame(foo=c(1, 2), bar=c("a", "b"))
rbind(df1, df2)

   foo bar
1   1   a
2   2   b

df1 <- data.frame(matrix(0, 0, 2))
names(df1) <- names(df2)
rbind(df1, df2)

   foo bar
1   1   a
2   2   b

----------------------------------------------
David L Carlson
Associate Professor of Anthropology
Texas A&M University
College Station, TX 77843-4352

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
project.org] On Behalf Of Dmitrii I.
Sent: Wednesday, January 02, 2013 2:09 PM
To: r-help at r-project.org
Subject: [R] rbind: inconsistent behaviour with empty data frames?

The rbind on empty and nonempty data frames behaves inconsistently. I
am
not sure if by design.

In the first example, first row is deleted, which may or may not be on
purpose:
df1 <- data.frame()
df2 <- data.frame(foo=c(1, 2), bar=c("a", "b"))
rbind(df1, df2)
     foo  bar
2   2     b

Now if we continue:
df1 <- data.frame(matrix(0, 0, 2))
names(df1) <- names(df2)
     foo bar
2     1   a
3     2   b
So now data frames combined successfully, but the row names have been
increased by one.

Finally, if row names are strings, it works okay:
row.names(df2) <- c("row1", "row2")
rbind(df1, df2)
           foo bar
row1    1   a
row2    2   b

Is this behaviour by design?

	[[alternative HTML version deleted]]