[UFO Chicago] gnuplot is love

[Brian Sobolak]

On Fri, August 5, 2011 7:37 am, Neil R. Ormos wrote:
> Brian Sobolak wrote:
>
>> Some of you may recall my search for a better
>> graphical program some time ago. In a paper I
>> read recently (which I'll post about separately)
>> I found a reference to gnuplot.
>> (www.gnuplot.info)
>
>> It's great.  I really like it, and would
>> recommend it.  It reminds me of the simplicity
>> of command-line FTP -- simple syntax,
>> scriptable, and has a lightweight GUI that I can
>> use to experiment.  It seems like a pretty fun
>> way to explore data, which I expect to do more
>> of in the coming months and am looking for some
>> new toolkits.
>
> Have you given up on R?

No, I don't think so.  I hadn't realized that O'Reilly has come out with
some books about R since the last time I was investigating this, and I
might go back and revisit it.

There's a subtle difference because gnuplot is really about just
visualizing/plotting data.  R would be better for actual statistical work,
ie categorization, standard deviation, etc.  gnuplot assumes that's
already been computed (*for the most part*) before you try to plot.

brian

ps At the next meeting, which will probably be at my house (more on that
in a bit), I can talk about some fun I'm having with python's eval()
function, which I was just delighted to learn that java doesn't have.

--