[R] Reading Multi-value data fields for descriptive analysis (resend)

From: Hohm, Dale <dale.hohm_at_hp.com>
Date: Sat, 12 Jul 2008 17:28:41 +0000


(I checked the archive and it seems the entire text of my post was 'scrubbed' - trying again)

Hello,

I'm looking for help on the best approach to get "multi-value" data fields into R for simple descriptive analysis.


I am new to this list and new to R, but I really want to get over the hump and get productive with it. Some help with how to best get the following data into R would be greatly appreciated. I have programming experience and stale experience with SPSS.

I am trying to do some simple descriptive analysis (frequencies, cross-tabs) of data stored in a Microsoft SharePoint list. The data can be accessed with ODBC or it can readily be extracted into an Excel or CSV format. One of the challenges with the data is that it uses several "multi-value" fields (Microsoft Access provides the same data-type).

By "multi-value" I mean that multiple responses are packed into a single data column; the data input form presents a question with several checkboxes and a free-format write-in response. The individual values within the data field are separated with the two characters ";#". So, the data would be of the following format (in CSV form with column headers and a tilde as the field separator):

Column1single~Column2multi~Column3multi
a sample value~C2 a multi one;#C2 a multi two~C3 a multi one;#C3 a multi two;#C3 a free-form answer

The first approach that comes to mind is to explode the multi-value fields into unique bi-variate data columns and then assign a 0 or 1 to these new columns in each record based on whether that specific value was present. This approach is complicated by the free-form answer as the unique columns could grow very large in number - it might be better to figure out how to indicate the presence of the free-form value in a data column called "Other" (or "C2 Other") and then hold the free-form value in a separate column.

The data would then look like this...

Column1single: a sample value
C2 a multi one: 1
C2 a multi two: 1
C2 a multi three: 0
C3 a multi one: 1
C3 a multi two: 1
C3 a free-form answer: 1
C3 another free-form answer: 0

Or in the second scenario...

Column1single: a sample value
C2 a multi one: 1
C2 a multi two: 1
C2 a multi three: 0
C3 a multi one: 1
C3 a multi two: 1
C3 Other: 1
C3 Other Text: a free-form answer

I am uncertain help to read this data into R in this format, so suggestions and examples would help me greatly.

This is a pretty common data packing scenario, so perhaps there are better approaches to reading this data and better ways in R to analyze it than what I have presented. Suggestions greatly appreciated.

Thanks,

Dale Hohm



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. Received on Sat 12 Jul 2008 - 17:34:06 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Sat 12 Jul 2008 - 18:31:17 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help. Please read the posting guide before posting to the list.

list of date sections of archive