[R] Novel (Maybe?) Visualizations

0 views
Skip to first unread message

D. Dashcle

unread,
Feb 9, 2010, 7:13:58 PM2/9/10
to r-h...@r-project.org

I'm interested in using R's plotting capabilities to try to generate a graph showing the relationship/pairing frequency of words appearing in a block of unstructured text. I don't have a specific algorithm or approach in mind, just looking to portray text in an interesting fashion. The output I'm hoping for is something like the following called a "Phrase Network":

http://manyeyes.alphaworks.ibm.com/manyeyes/page/Phrase_Net.html

I found this package which seems semi-related, but I haven't been getting great results and I'm not sure if I'm on the right track.

http://tm.r-forge.r-project.org/screenshots.html

Does anyone have suggestions for another package I should look into?

-D

_________________________________________________________________
Hotmail: Free, trusted and rich email service.

[[alternative HTML version deleted]]

______________________________________________
R-h...@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.

Johannes Huesing

unread,
Feb 10, 2010, 3:11:20 PM2/10/10
to r-h...@r-project.org
D. Dashcle <d.da...@hotmail.com> [Wed, Feb 10, 2010 at 01:13:58AM CET]:
>

> I'm interested in using R's plotting capabilities to try to generate
a graph showing the relationship/pairing frequency of words
appearing in a block of unstructured text. I don't have a specific
algorithm or approach in mind, just looking to portray text in an
interesting fashion. The output I'm hoping for is something like the
following called a "Phrase Network":
>
> http://manyeyes.alphaworks.ibm.com/manyeyes/page/Phrase_Net.html

It looks like you need algorithms to position nodes in order to minimise
edge lengths and overlap. A well known program which does that is
graphviz. I am quite sure that no one has taken the effort to re-implement
the algorithm in R, but there are many R packages that interact with graphviz.

I would look for "graphviz" in www.rseek.org and procede from there.

--
Johannes Hüsing There is something fascinating about science.
One gets such wholesale returns of conjecture
mailto:joha...@huesing.name from such a trifling investment of fact.
http://derwisch.wikidot.com (Mark Twain, "Life on the Mississippi")

stvienna wiener

unread,
Feb 12, 2010, 12:19:42 AM2/12/10
to joha...@huesing.name, r-h...@r-project.org
2010/2/10 Johannes Huesing <joha...@huesing.name>:

> D. Dashcle <d.da...@hotmail.com> [Wed, Feb 10, 2010 at 01:13:58AM CET]:

> I would look for "graphviz" in www.rseek.org and procede from there.

Is graphviz really the only option? I guss so, but the license of
graphviz is very restricted and not comparable
to the usal open source licenses, e.g. GPL or similar.


I spent about a day to find information on graphs in R, search and
searched, tried many libraries, etc.
Compared to other subjects or areas, I found very little and the
quality was limited.

What I need would be:

- import/export from GraphML
- basic and advanced graph algorithms

- nice visualization of graphs (graphviz)
- interactive input of graphs (or external program and import via GraphML)


Is R the right programming language for that? (I think it is...)

Are there any pointers? Books or tutorials on using R for graph stuff.?

Any help is very much appreciated!

Best regards,
St.

Liviu Andronic

unread,
Feb 12, 2010, 4:54:45 AM2/12/10
to stvienna wiener, r-h...@r-project.org
On Fri, Feb 12, 2010 at 5:19 AM, stvienna wiener <stvi...@gmail.com> wrote:
>   - interactive input of graphs (or external program and import via GraphML)
>
Perhaps latticist, rggobi?

> Is R the right programming language for that? (I think it is...)
>
> Are there any pointers? Books or tutorials on using R for graph stuff.?
>

This [1][2] might provide some pointers. Also look at the Graphs Task View.
Liviu

[1] http://www.metaresearch.de/exlib/
[2] http://bm2.genes.nig.ac.jp/RGM2/index.php?clear=all

Reply all
Reply to author
Forward
0 new messages