xsequence - a grid-changing function

[Elizabeth Dobbins <dobbins@xxxxxxxxxxxxx>]

Dear Ferreters;

I wrote last week about a question I had with xsequence.  When I
investigated further, I found the problem was more complicated than I
originally thought.  Namely, when the region was specified in the command
line, xsequence worked predictably, but when I used 'set region' or
'repeat' to define the region, xsequence returned what I thought were
unpredictable results.  For instance,

yes? stat xsequence(sst[x=20E:30E,y=0N:20N,l=1])

and

yes? set region/x=20E:30E/y=0N:20N/l=1
yes? stat xsequence(sst)

return 2 very different answers.

According to Steve, this is a feature of a new class of ferret functions:
grid-changing functions.  For grid-changing functions, the axes of the
result may differ from the axes of the arguments.  Therefore, if you want
to specify a region, you must be very precise about whether you are
referring to the region of the argument, or the region of the result.  In
my example above, the first command specifies the region of the argument,
and the second specifies the region of the result.  

This subtlety was not described well in the User Guide, but a FAQ will
soon be added to the ferret home page that will discuss the issue in more
detail.

Thanks to all that responded.
Liz


*********************
Elizabeth L. Dobbins
Research Scientist
JISAO/PMEL
phone: (206) 526-4581

Fine words butter no parsnips.