R help - Sep 2007 - reshape() to wide with varying number of responses to fields

Hello,

I have a dataframe containing the following variables:

	PID, Field, Value

Where PID refers to a unique individual, Field to a particular question, and
Value to a particular response to a question.

I?d like this in wide format, with a different row for each PID.
However.... there are differing numbers of Values associated with each
Field, for each PID. 
For example, a Field question may be ?Which of the following are important
for you:? ? some people tick one box in response to it, other 3 or 4: thus
PID A might generate 3 Values for Field X, whereas PID B would only generate
1 Value:

PID	Field Value
A	X	a
A	X	b
A	X	c
B	X	b
Etc. 
What I?d like is a wide format, with one Pid per row, and enough columns for
each of the possible values the PID could provide for a particular field,
filled in with the values if they exist, or NAs if they don?t:

PID	X.1	X.2	X.3
A	a	b	c
B	b	NA	NA
Etc. 

I already know the maximum number of Values any PID has provided for each
Field (i.e. the maximum number of columns to allow for each Field response),
but don?t know how to create this dataframe using reshape (or if reshape is
the right function to use). 


Any help much appreciated. I?ve not seen a ?how-to? regarding this in the
archive.
Thanks, Jon




Checked by AVG Free Edition. 

20/09/2007
12:07

Hi Jon,

You can do this with the reshape package.  Your data are already in
"molten" form, so you can just do:

cast(rename(df, c("Value" = "value"), PID ~ Field)

You can find out more about reshape at http://had.co.nz/reshape

Hadley

On 9/22/07, Jon Minton <jm540 at york.ac.uk>
wrote:
> Hello,
>
> I have a dataframe containing the following variables:
>
>         PID, Field, Value
>
> Where PID refers to a unique individual, Field to a particular question,
and
> Value to a particular response to a question.
>
> I'd like this in wide format, with a different row for each PID.
> However.... there are differing numbers of Values associated with each
> Field, for each PID.
> For example, a Field question may be "Which of the following are
important
> for you:" ? some people tick one box in response to it, other 3 or 4:
thus
> PID A might generate 3 Values for Field X, whereas PID B would only
generate
> 1 Value:
>
> PID     Field Value
> A       X       a
> A       X       b
> A       X       c
> B       X       b
> Etc.
> What I'd like is a wide format, with one Pid per row, and enough
columns for
> each of the possible values the PID could provide for a particular field,
> filled in with the values if they exist, or NAs if they don't:
>
> PID     X.1     X.2     X.3
> A       a       b       c
> B       b       NA      NA
> Etc.
>
> I already know the maximum number of Values any PID has provided for each
> Field (i.e. the maximum number of columns to allow for each Field
response),
> but don't know how to create this dataframe using reshape (or if
reshape is
> the right function to use).
>
>
> Any help much appreciated. I've not seen a 'how-to' regarding
this in the
> archive.
> Thanks, Jon
>
>
>
>
> Checked by AVG Free Edition.
>
> 20/09/2007
> 12:07
>
>
> ______________________________________________
> 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.
>

-- 
http://had.co.nz/