Re: PLOT procedure (was: Julian Day Numbers)

[Martin Schultz <martin.schultz(at)dkrz.de>]

Craig Markwardt wrote:
> 
> James Tappin <sjt@star.sr.bham.ac.uk> writes:
> 
> > One real gotcha with Julian day numbers is if you try to use them for a
> > time axis on a plot.
> >
> > It works fine for long time axes, but if you try looking at high-resolution
> > data things look really messy.
> >
> > The problem is that PLOT converts to single-precision before converting
> > coordinates to data and this means that all values are truncated to the
> > nearest quarter day (for IEEE floats).
> 
> I agree that double precision plotting is important.  Sometimes you
> just need it.  However I think that in this case you really want to
> subtract a time offset anyway.  Consider that your axis labels will
> never be pretty if you are using full Julian days.  [ for those that
> don't know Julian days of this millenium number in the millions ].
> You can put in your XTITLE or TITLE the time offset and be done with
> it.
> 
> Nit-pickily yours,
> Craig
> 
> --
> --------------------------------------------------------------------------
> Craig B. Markwardt, Ph.D.         EMAIL:    craigmnet@cow.physics.wisc.edu
> Astrophysics, IDL, Finance, Derivatives | Remove "net" for better response
> --------------------------------------------------------------------------

... but then again: what if you use the tickformat keyword with the
label_date function? I recently gave this a try and had to be careful
with roundoff errors (in IDL5.3).

Martin

-- 
[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[
[[ Dr. Martin Schultz   Max-Planck-Institut fuer Meteorologie    [[
[[                      Bundesstr. 55, 20146 Hamburg             [[
[[                      phone: +49 40 41173-308                  [[
[[                      fax:   +49 40 41173-298                  [[
[[ martin.schultz@dkrz.de                                        [[
[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[