R help - Nov 2010 - Extracting from data.frame

Can someone please explain the following behavior?

 

df1 and df2 are data.frames. Suppose I want a subset of the rows
(observations) using extraction, say just the first row. What I want to know
is why if df1 has just one column then df1[1,] returns a vector, whereas if
df2 has 2 or more columns then df2[1,] returns a data frame? Why is the
single column case different?


	[[alternative HTML version deleted]]

Use
   df[1,,drop=FALSE]
if you want the output to be a data.frame,
no matter how many columns df has.

The effective default value for drop
is TRUE if the output has a single column.
This does not depend on whether or not
you subscripted rows or columns or both.

Bill Dunlap
Spotfire, TIBCO Software
wdunlap tibco.com  
> -----Original Message-----
> From: r-help-bounces at r-project.org 
> [mailto:r-help-bounces at r-project.org] On Behalf Of James Hirschorn
> Sent: Sunday, November 14, 2010 4:32 PM
> To: R-help at r-project.org
> Subject: [R] Extracting from data.frame
> 
> Can someone please explain the following behavior?
> 
>  
> 
> df1 and df2 are data.frames. Suppose I want a subset of the rows
> (observations) using extraction, say just the first row. What 
> I want to know
> is why if df1 has just one column then df1[1,] returns a 
> vector, whereas if
> df2 has 2 or more columns then df2[1,] returns a data frame? 
> Why is the
> single column case different?
> 
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

It's because in this instance R (and S before it) behaves a bit like
Microsoft and tries to guess what you really wanted rather than listen to what
you asked for.

If you ask for a single column from a data frame, e.g.

df[,1]

then by default this behaves like df[[1]] and you get a single component, no
longer a part of the data frame.  If you want to ensure you always get a data
frame, you need to say so with

df[,j, drop = FALSE]

If you select a single row from a data frame in general you get a data frame. 
The exceptin is when the data frame itself only has one column and the dropping
convention swings into action once more.  The way to avoid this is the same,
i.e.

df[j,,drop = FALSE]

In general if you want to ensure consistent behaviour in scripts, it is a
recommended that you use "drop = FALSE" as a kind of pseudo-index in
the last position like this.  You can always demote the object to a simpler
form, if that's what you really want and if one exists, using the drop()
function later.

Why is it so?  It was not my idea.  I do know, though, that the Matrix package
allows matrix operations with this convention inoperative.  Other people must
have become a tad frustrated with it before now.

Bill Venables. 

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
On Behalf Of James Hirschorn
Sent: Monday, 15 November 2010 10:32 AM
To: R-help at r-project.org
Subject: [R] Extracting from data.frame

Can someone please explain the following behavior?

 

df1 and df2 are data.frames. Suppose I want a subset of the rows
(observations) using extraction, say just the first row. What I want to know
is why if df1 has just one column then df1[1,] returns a vector, whereas if
df2 has 2 or more columns then df2[1,] returns a data frame? Why is the
single column case different?


	[[alternative HTML version deleted]]

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.