| Re: Why does IDL strip unary dimensions from structure elements? [message #5236] |
Fri, 27 October 1995 00:00 |
steinhh
Messages: 260
Registered: June 1994
|
Senior Member |
|
|
In article <308FD223.167E@grossc.gsfc.nasa.gov>, Thomas A McGlynn <mcglynn@grossc.gsfc.nasa.gov> writes:
The problem with trailing singular array dimensions disappearing:
|> Even this is not guaranteed:
|>
|> a={e:intarr(1,1,1)}
|> help,a,/str yields:
|>
|> ** Structure <40014f08>, 1 tags, length=2, refs=1:
|> E INT Array(1)
|> I.e., it got rid of some of the dimensions but not all. Nor can I use
|> reform to fix a structure element. It's a mess. Has anyone mentioned
|> this to RSI?
|>
I tried to start a discussion on this a couple of months(?) ago, but
there wasn't much response. Probably because people seldom come across
the problem in everyday use. Once you notice it, though, it's terribly
bothersome, adding a great deal of complexity to the otherwise very
simple array operations. Since we have a simple way of removing (all)
singular dimensions (REFORM), why can't we keep them unless told
otherwise!
There's one "good" thing about it: The possibility of indexing
everything with an extra zero does compensate a little for the
other "feature".
If we have, e.g.,
a=intarr(10,1,1)
help,a
a INT = Array(10)
Then the following will not be an error, and it will produce correct
results:
element = a(10,0,0)
So long as you're indexing your array with legal values for the indices
(i.e., zero for all singular dimensions), then you get what you expect.
It's "just" a matter of not letting your program be confused about it :-)
Stein Vidar
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
