| Re: Maddening structures [message #50209] |
Mon, 18 September 2006 09:40 |
Jean H.
Messages: 472
Registered: July 2006
|
Senior Member |
|
|
> Fine, unless there might be further tags (which you can then solve
> with a bit of looping and executing), and complicated when they aren't
> all floats but a mix of floats, ints, longs and strings. See my other
> post...
What about using an array of pointers? drop the structure and work on
the array! ... it could be of any size!
Jean
|
|
|
|
| Re: Maddening structures [message #50218 is a reply to message #50209] |
Mon, 18 September 2006 07:16  |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Edd writes:
> Think I'm getting there though. Whether anyone will be able to manage
> this code later is another question.
Just leave all the comments out. Then, whoever has to
deal with it later will have to rewrite it the way
it SHOULD have been written in the first place. :-)
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Sepore ma de ni thui. ("Perhaps thou speakest truth.")
|
|
|
|
| Re: Maddening structures [message #50219 is a reply to message #50218] |
Mon, 18 September 2006 06:55 |
Edd Edmondson
Messages: 50
Registered: January 2003
|
Member |
|
|
David Fanning <davidf@dfanning.com> wrote:
> Edd writes:
>> OK, I've got some data read in so that I have a structure that looks
>> like
>>
>> data.foo[600]
>> .bar[600]
>> .baz[600]
>>
>> and I want
>> data[600].foo
>> .bar
>> .baz
>>
>> In other words I want an array of structures rather than a structure
>> of arrays.
>>
>> Does anyone have any magic that works for the case when foo, bar and
>> baz are not previously known?
> names = ['foo','bar', 'baz']
> struct = Create_Struct(names[0], 0.0, names[1], 0.0, names[2], 0.0)
> data = Replicate(struct, 600)
Fine, unless there might be further tags (which you can then solve
with a bit of looping and executing), and complicated when they aren't
all floats but a mix of floats, ints, longs and strings. See my other
post...
Think I'm getting there though. Whether anyone will be able to manage
this code later is another question.
--
Edd
|
|
|
|
| Re: Maddening structures [message #50220 is a reply to message #50219] |
Mon, 18 September 2006 06:54 |
Edd Edmondson
Messages: 50
Registered: January 2003
|
Member |
|
|
Edd <eddedmondson@hotmail.com> wrote:
> That's where I was looking too. I wasn't sure how to handle the
> structure creation neatly though. Once you have the tag names you can
> loop over them in a for loop (there won't be that many tags anyway) so
> it's not too bad. Just a bit hairy. It'd be nice if there was a
> reform-alike. :-)
Saying it's not too bad is an exaggeration. All this is happening
inside an execute(), so I'm entering it in one set of ''s. I need to
put it all in another execute() inside that to make it
on-the-fly-enough, which uses up a "" inside the '', leaving me using
STRING(39B) to get a third tier. I've managed to build up a string
that looks like
'mystructure=create_struct('foo',0.1,'bar',1242999,'baz',som estring)'
which means I get my float in foo, my long in bar, but baz complains
because the string is coming out without the ''s it needs, only I
can't put those ''s in in general or the floats and longs turn into
strings too, breaking other stuff.
It's the ugliest code I've written (well, since that bit of Perl I did
last week) and I'm about to throw a brick at the screen. All this just
to rejig a structure a bit. Surely this can be done with histogram()
or something ;-)
--
Edd
|
|
|
|
| Re: Maddening structures [message #50221 is a reply to message #50220] |
Mon, 18 September 2006 06:52 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Edd writes:
> OK, I've got some data read in so that I have a structure that looks
> like
>
> data.foo[600]
> .bar[600]
> .baz[600]
>
> and I want
> data[600].foo
> .bar
> .baz
>
> In other words I want an array of structures rather than a structure
> of arrays.
>
> Does anyone have any magic that works for the case when foo, bar and
> baz are not previously known?
names = ['foo','bar', 'baz']
struct = Create_Struct(names[0], 0.0, names[1], 0.0, names[2], 0.0)
data = Replicate(struct, 600)
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Sepore ma de ni thui. ("Perhaps thou speakest truth.")
|
|
|
|
| Re: Maddening structures [message #50222 is a reply to message #50221] |
Mon, 18 September 2006 05:24 |
Edd Edmondson
Messages: 50
Registered: January 2003
|
Member |
|
|
Maarten <maarten.sneep@knmi.nl> wrote:
> Edd wrote:
>> OK, I've got some data read in so that I have a structure that looks
>> like
>>
>> data.foo[600]
>> .bar[600]
>> .baz[600]
>>
>> and I want
>> data[600].foo
>> .bar
>> .baz
>>
>> In other words I want an array of structures rather than a structure
>> of arrays.
>>
>> Does anyone have any magic that works for the case when foo, bar and
>> baz are not previously known?
> Perhaps my reading comprehension isn't quite awake yet, but:
> 1) do you mean the names themselves, or the number of fields (there may
> also be a fuu around, in addition to the foo, bar and baz)?
I'd basically like the most general method I can find (so I don't have
to go recoding stuff when another field gets added a week or two down
the line). So number and naming of fields should be considered flexible.
> Get the names with tag_names(data), obtain the length with
> size(data.(0), /dim), create a base struct with base =
> CREATE_STRUCT(...) and replicate() the base to the right size. At least
> that is where I would look. Copying is another matter, perhaps others
> can comment on how to do this efficiently. At least you have the right
> structure to store the stuff.
That's where I was looking too. I wasn't sure how to handle the
structure creation neatly though. Once you have the tag names you can
loop over them in a for loop (there won't be that many tags anyway) so
it's not too bad. Just a bit hairy. It'd be nice if there was a
reform-alike. :-)
I may try for an alternative workaround, as rejigging the other bits
probably involves less recoding than doing the jigging with this.
--
Edd
|
|
|
|
| Re: Maddening structures [message #50223 is a reply to message #50222] |
Mon, 18 September 2006 05:23 |
Edd Edmondson
Messages: 50
Registered: January 2003
|
Member |
|
|
Maarten <maarten.sneep@knmi.nl> wrote:
> Edd wrote:
>> OK, I've got some data read in so that I have a structure that looks
>> like
>>
>> data.foo[600]
>> .bar[600]
>> .baz[600]
>>
>> and I want
>> data[600].foo
>> .bar
>> .baz
>>
>> In other words I want an array of structures rather than a structure
>> of arrays.
>>
>> Does anyone have any magic that works for the case when foo, bar and
>> baz are not previously known?
> Perhaps my reading comprehension isn't quite awake yet, but:
> 1) do you mean the names themselves, or the number of fields (there may
> also be a fuu around, in addition to the foo, bar and baz)?
I'd basically like the most general method I can find (so I don't have
to go recoding stuff when another field gets added a week or two down
the line). So number and naming of fields should be considered flexible.
> Get the names with tag_names(data), obtain the length with
> size(data.(0), /dim), create a base struct with base =
> CREATE_STRUCT(...) and replicate() the base to the right size. At least
> that is where I would look. Copying is another matter, perhaps others
> can comment on how to do this efficiently. At least you have the right
> structure to store the stuff.
That's where I was looking too. I wasn't sure how to handle the
structure creation nearly though. Once you have the tag names you can
loop over them in a for loop (there won't be that many tags anyway) so
it's not too bad. Just a bit hairy. It'd be nice if there was a
reform-alike. :-)
I may try for an alternative workaround, as rejigging the other bits
probably involves less recoding than doing the jigging with this.
--
Edd
|
|
|
|
| Re: Maddening structures [message #50224 is a reply to message #50223] |
Mon, 18 September 2006 05:13 |
Maarten[1]
Messages: 176
Registered: November 2005
|
Senior Member |
|
|
Edd wrote:
> OK, I've got some data read in so that I have a structure that looks
> like
>
> data.foo[600]
> .bar[600]
> .baz[600]
>
> and I want
> data[600].foo
> .bar
> .baz
>
> In other words I want an array of structures rather than a structure
> of arrays.
>
> Does anyone have any magic that works for the case when foo, bar and
> baz are not previously known?
Perhaps my reading comprehension isn't quite awake yet, but:
1) do you mean the names themselves, or the number of fields (there may
also be a fuu around, in addition to the foo, bar and baz)?
Get the names with tag_names(data), obtain the length with
size(data.(0), /dim), create a base struct with base =
CREATE_STRUCT(...) and replicate() the base to the right size. At least
that is where I would look. Copying is another matter, perhaps others
can comment on how to do this efficiently. At least you have the right
structure to store the stuff.
Maarten
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
