[MATRIX-SIG] making a text label from an object instance?

[Phil Austin]

>>>>> "Andrew" == Andrew P Mullhaupt <amullhau@ix.netcom.com> writes:

    Andrew> At 11:43 AM 7/10/97 +0200, Konrad Hinsen wrote:
    >> I wouldn't worry too much about default labels. I find them
    >> rarely useful. It's simple enough to pass an explicit string.

    Andrew> Don't do a lot of automatically generated plots, eh?

    Andrew> Default labels are _very_ useful. I don't see a really
    Andrew> easy way to do it in python, but it's one of the things
    Andrew> people really like when it's available.


I'd agree with this--looking at a suprising result in your notebook
several weeks after the plot was produced, it's very reassuring to
know that there is no possible way the axis labels could have been
swapped.  The variables in the label can be made more informative
using named dimensions and range indexes, so that contours of a 4-d
array of vertical velocity don't really need a title:

contour(vertical_velocity["500milibars","300Kisotherm",lats,lons])

combined with a time-stamp that can be printed on the plot and used in
an RCS comment when you check in the plotting script, it's possible to
create 500 fully documented data plots without any programmer
intervention, and recover any one of them six months later from the
raw netcdf data after your grad student has spilled root beer
on the original.

There's no looking back though.  Splus has the treasured
"deparse(substitute())" but they've doubled their site licence fees
and seem to be preparing to abandon Unix as a supported platform.
Brian Ripley (http://lib.stat.cmu.edu/S/MASS2/), for example, recently
posted a message in S-news that sounded positively bitter.  Picking up
some Splus refugees who happen to be world-renown in
bootstrap/LOESS/ARIMA techniques would be major coup for Python
(albeit a fairly distant prospect).

Regards, Phil




_______________
MATRIX-SIG  - SIG on Matrix Math for Python

send messages to: matrix-sig@python.org
administrivia to: matrix-sig-request@python.org
_______________