[Met_help] [rt.rap.ucar.edu #62536] History for Query Regarding plot_cnt.R (Rscript)

[John Halley Gotway via RT]

----------------------------------------------------------------
  Initial Request
----------------------------------------------------------------

Hello John,

My query is regarding plot_cnt.R is plotting single models output, right?
What if I've have multiple models output? I know it can make plots for
multiple models output, in separated/ attached plot.

Actually I'm new to Rscript and I had tried to modify your scripts, but
couldn't succeed.
Now what I want, I want a little modification in the plot_cnt.R scripts
which can plot the multiple models output( here  I am using  GFS, UM, and
SAM [ South  Asian Model] ) to the single plot. e.g. If I am considering
RMSE, for Box & Whiskers plot and time series plot it can give in different
colors (Legend) or different type of lines in a single plot. Is it
possible? If so please help me.

Thank you for your valuable time.

Waiting for your reply.
-- 
*     Regards*
===========================================
      HARVIR SINGH
       Project Scientist- 'C'
       *National Centre for Medium Range Weather Forecasting
*
*       (Ministry of Earth Sciences)
*
*       NOIDA Sector-62 (INDIA)*
       harvir at ncmrwf.gov.in,
       Mobile No. +91-9456279936, extn. 297
       http://www.harvir007.webs.com
===========================================


----------------------------------------------------------------
  Complete Ticket History
----------------------------------------------------------------

Subject: Re: [rt.rap.ucar.edu #62536] Query Regarding plot_cnt.R (Rscript)
From: John Halley Gotway
Time: Thu Aug 08 10:04:10 2013

Harvir,

We provide the plot_cnt.R script as a very basic, quick way to
visualize some MET output.  We do not intend it to be robust plotting
system.  We do have such a system that we use internally, named
METViewer.  However, we don't currently have adequate funding to
support it to the user community.  We are exploring ways to make
METViewer available to users in the future.

So at this point, I'd suggest using whatever plotting package you're
most comfortable with.  The MET output is just columnar ASCII data -
and there are many scripting/plotting tools that can easily
read that sort of data.  We use R heavily in the DTC, but you could
easily plot the MET output using matlab, IDL, or even microsoft excel.

When comparing the output of two models, one big consideration is
event equalization.  You want to make sure that you're comparing their
performance on the exact same set of cases.

If you're interested in learning more about R, here's some good
introductory information:
    http://cran.r-project.org/doc/manuals/R-intro.html

But providing modifications to the sample plotting scripts is beyond
the scope of support through MET Help.

Thanks,
John Halley Gotway
met_help at ucar.edu



On 08/06/2013 10:45 PM, HARVIR SINGH via RT wrote:
>
> Tue Aug 06 22:45:22 2013: Request 62536 was acted upon.
> Transaction: Ticket created by harviriitkgp at gmail.com
>         Queue: met_help
>       Subject: Query Regarding plot_cnt.R (Rscript)
>         Owner: Nobody
>    Requestors: harviriitkgp at gmail.com
>        Status: new
>   Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=62536 >
>
>
> Hello John,
>
> My query is regarding plot_cnt.R is plotting single models output,
right?
> What if I've have multiple models output? I know it can make plots
for
> multiple models output, in separated/ attached plot.
>
> Actually I'm new to Rscript and I had tried to modify your scripts,
but
> couldn't succeed.
> Now what I want, I want a little modification in the plot_cnt.R
scripts
> which can plot the multiple models output( here  I am using  GFS,
UM, and
> SAM [ South  Asian Model] ) to the single plot. e.g. If I am
considering
> RMSE, for Box & Whiskers plot and time series plot it can give in
different
> colors (Legend) or different type of lines in a single plot. Is it
> possible? If so please help me.
>
> Thank you for your valuable time.
>
> Waiting for your reply.
>

------------------------------------------------