| Re: findgen([variable]) [message #22890] |
Tue, 12 December 2000 11:50 |
Pavel A. Romashkin
Messages: 531
Registered: November 2000
|
Senior Member |
|
|
Not only was my suggestion "Oof", but it was downright wrong, too, I am
afraid :-( Perfect mental state to get ready for that conference, Pavel.
But, if I didn't post this ugly one, how else would this thread stir up
so much nice input from the better ones? You guys were dormant there for
a couple of hours!
Cheers,
Pavel
Paul van Delst wrote:
> Oof!
> ...... snip
> paulv
|
|
|
|
| Re: findgen([variable]) [message #22891 is a reply to message #22890] |
Tue, 12 December 2000 11:33  |
John-David T. Smith
Messages: 384
Registered: January 2000
|
Senior Member |
|
|
Craig Markwardt wrote:
>
> JD Smith <jdsmith@astro.cornell.edu> writes:
>
>> Craig Markwardt wrote:
>>>
>>> Martin Skou Andersen <skou@fys.ku.dk> writes:
>>>> Is it posible to make an mutiple dimensional array by using a variable
>>>> such as findgen(x)?
>>>
>>> This is virtually impossible, right? You are asking for each row of
>>> the resulting array to be of a different length. That's not really a
>>> 2 dimensional array is it?
>>
>>
>> Are we sure he doesn't just mean:
>>
>> findgen(x[0],x[1]) ?
>
> Ah yes, that would be a good idea. In fact, one of my "top ten"
> requests was to have all of the functions that accept dimensions,
> accept them in a consistent way. REBIN, REFORM and TOTAL are all
> different. We should be able to specify dimensions consistently
> between them, and either as an array, or a list of scalars.
Or as either, as in Perl, where
@a=($a1,$a2);
sub(@a);
sub($a1,$a2)
are the same. A list is a list is a list. This simplifies a lot of
things (and of course requires atomic list members (i.e. pointers),
since you can't have separate lists and scalars all on the command
line).
Not that we could ever get away with that kind of facelift, but PerlDL
of course has hope.
JD
|
|
|
|
| Re: findgen([variable]) [message #22892 is a reply to message #22891] |
Tue, 12 December 2000 11:24 |
Paul van Delst
Messages: 364
Registered: March 1997
|
Senior Member |
|
|
Craig Markwardt wrote:
>
> Paul van Delst <pvandelst@ncep.noaa.gov> writes:
>> Oof!
>
> Looks like we posted at the same time. Oof indeed. :-)
>
>> Why doesn't MAKE_ARRAY allow for a dimension vector input?
>
> It does, ever try the DIMENSION keyword, sonny? :-)
Oh yeah.... I still haven't received your post on my server (you're at GSFC, right, so that's
probably why...last on the list of NOAA server updates :o) so this is probably the same as you put
together:
IDL> x=[5,7]
IDL> help, make_array(dimension=x,/index)
<Expression> FLOAT = Array[5, 7]
IDL> print, make_array(dimension=x,/index)
0.00000 1.00000 2.00000 3.00000 4.00000
5.00000 6.00000 7.00000 8.00000 9.00000
10.0000 11.0000 12.0000 13.0000 14.0000
15.0000 16.0000 17.0000 18.0000 19.0000
20.0000 21.0000 22.0000 23.0000 24.0000
25.0000 26.0000 27.0000 28.0000 29.0000
30.0000 31.0000 32.0000 33.0000 34.0000
Veddy nice
No need for any "PRODUCT"-ing
cool
paulv
--
Paul van Delst Ph: (301) 763-8000 x7274
CIMSS @ NOAA/NCEP Fax: (301) 763-8545
Rm.207, 5200 Auth Rd. Email: pvandelst@ncep.noaa.gov
Camp Springs MD 20746
|
|
|
|
| Re: findgen([variable]) [message #22893 is a reply to message #22892] |
Tue, 12 December 2000 11:07 |
Craig Markwardt
Messages: 1869
Registered: November 1996
|
Senior Member |
|
|
Paul van Delst <pvandelst@ncep.noaa.gov> writes:
> Oof!
Looks like we posted at the same time. Oof indeed. :-)
> Seems that determining "n" is the hard part...is there a "PRODUCT" type of function in IDL? Like
> TOTAL but with * rather than + as the operator.
There is PRODUCT in the IDL Astronomy Library, and CMPRODUCT from my
web page, which I purport to be faster in most cases. Alas, there is
no built-in IDL support, but there should be. Make that my "top
eleventh" request.
> Why doesn't MAKE_ARRAY allow for a dimension vector input?
It does, ever try the DIMENSION keyword, sonny? :-)
Craig
http://cow.physics.wisc.edu/~craigm/idl/idl.html
--
------------------------------------------------------------ --------------
Craig B. Markwardt, Ph.D. EMAIL: craigmnet@cow.physics.wisc.edu
Astrophysics, IDL, Finance, Derivatives | Remove "net" for better response
------------------------------------------------------------ --------------
|
|
|
|
| Re: findgen([variable]) [message #22894 is a reply to message #22893] |
Tue, 12 December 2000 10:52 |
Craig Markwardt
Messages: 1869
Registered: November 1996
|
Senior Member |
|
|
JD Smith <jdsmith@astro.cornell.edu> writes:
> Craig Markwardt wrote:
>>
>> Martin Skou Andersen <skou@fys.ku.dk> writes:
>>> Is it posible to make an mutiple dimensional array by using a variable
>>> such as findgen(x)?
>>
>> This is virtually impossible, right? You are asking for each row of
>> the resulting array to be of a different length. That's not really a
>> 2 dimensional array is it?
>
>
> Are we sure he doesn't just mean:
>
> findgen(x[0],x[1]) ?
Ah yes, that would be a good idea. In fact, one of my "top ten"
requests was to have all of the functions that accept dimensions,
accept them in a consistent way. REBIN, REFORM and TOTAL are all
different. We should be able to specify dimensions consistently
between them, and either as an array, or a list of scalars.
Here is my entry. Sorry, no HISTOGRAM call, but I need to give it a
rest once in a while.
nel = dims(0)
for i = 1L, n_elements(dims)-1 do nel = nel * dims(i)
arr = reform(findgen(nel), dims)
Craig
--
------------------------------------------------------------ --------------
Craig B. Markwardt, Ph.D. EMAIL: craigmnet@cow.physics.wisc.edu
Astrophysics, IDL, Finance, Derivatives | Remove "net" for better response
------------------------------------------------------------ --------------
|
|
|
|
| Re: findgen([variable]) [message #22895 is a reply to message #22894] |
Tue, 12 December 2000 10:33 |
Paul van Delst
Messages: 364
Registered: March 1997
|
Senior Member |
|
|
"Pavel A. Romashkin" wrote:
>
> I am afraid FINDGEN can not take a vector for dimensions. What I would
> try is either using a vector
>
> x = [5, 7]
> r = findgen(total(x))
>
> which still would allow transparent indexing of R. If I really needed
> multi D arrays, I'd use EXECUTE and build a command string using
> dimensions of X:
>
> x = [5, 7]
> com = 'r = findgen('
> for i =0, n_elements(x)-1 do com = com +'x['+string(i)+']'+ $
> string(44b*byte(i ne n_elements(x)-1) > 41b)
> i=execute(com)
>
> EXECUTE is not cute, but I can't come up with anything nicer. I need
> another cup of coffe :-(
Oof!
What about
IDL> x=[5,7]
IDL> ndim=n_elements(x)
IDL> n=1L
IDL> for i=0,ndim-1 do n=n*x[i]
IDL> help, reform(findgen(n),x)
<Expression> FLOAT = Array[5, 7]
IDL> print, reform(findgen(n),x)
0.00000 1.00000 2.00000 3.00000 4.00000
5.00000 6.00000 7.00000 8.00000 9.00000
10.0000 11.0000 12.0000 13.0000 14.0000
15.0000 16.0000 17.0000 18.0000 19.0000
20.0000 21.0000 22.0000 23.0000 24.0000
25.0000 26.0000 27.0000 28.0000 29.0000
30.0000 31.0000 32.0000 33.0000 34.0000
Seems that determining "n" is the hard part...is there a "PRODUCT" type of function in IDL? Like
TOTAL but with * rather than + as the operator.
Why doesn't MAKE_ARRAY allow for a dimension vector input?
paulv
--
Paul van Delst Ph: (301) 763-8000 x7274
CIMSS @ NOAA/NCEP Fax: (301) 763-8545
Rm.207, 5200 Auth Rd. Email: pvandelst@ncep.noaa.gov
Camp Springs MD 20746
|
|
|
|
| Re: findgen([variable]) [message #22896 is a reply to message #22895] |
Tue, 12 December 2000 10:20 |
John-David T. Smith
Messages: 384
Registered: January 2000
|
Senior Member |
|
|
Craig Markwardt wrote:
>
> Martin Skou Andersen <skou@fys.ku.dk> writes:
>> Is it posible to make an mutiple dimensional array by using a variable
>> such as findgen(x)?
>
> This is virtually impossible, right? You are asking for each row of
> the resulting array to be of a different length. That's not really a
> 2 dimensional array is it?
Are we sure he doesn't just mean:
findgen(x[0],x[1]) ?
JD
|
|
|
|
| Re: findgen([variable]) [message #22899 is a reply to message #22896] |
Tue, 12 December 2000 09:53 |
Pavel A. Romashkin
Messages: 531
Registered: November 2000
|
Senior Member |
|
|
Oh Craig, I expected an answer involving HISTOGRAM from you :-(
Cheers,
Pavel
Craig Markwardt wrote:
>
> Martin Skou Andersen <skou@fys.ku.dk> writes:
>> Is it posible to make an mutiple dimensional array by using a variable
>> such as findgen(x)?
>
> This is virtually impossible, right? You are asking for each row of
> the resulting array to be of a different length. That's not really a
> 2 dimensional array is it?
>
> Craig
|
|
|
|
| Re: findgen([variable]) [message #22900 is a reply to message #22899] |
Tue, 12 December 2000 09:45 |
Pavel A. Romashkin
Messages: 531
Registered: November 2000
|
Senior Member |
|
|
I am afraid FINDGEN can not take a vector for dimensions. What I would
try is either using a vector
x = [5, 7]
r = findgen(total(x))
which still would allow transparent indexing of R. If I really needed
multi D arrays, I'd use EXECUTE and build a command string using
dimensions of X:
x = [5, 7]
com = 'r = findgen('
for i =0, n_elements(x)-1 do com = com +'x['+string(i)+']'+ $
string(44b*byte(i ne n_elements(x)-1) > 41b)
i=execute(com)
EXECUTE is not cute, but I can't come up with anything nicer. I need
another cup of coffe :-(
Cheers,
Pavel
Martin Skou Andersen wrote:
>
> Hi...
> I was wondering if I could use findgen with a variable.
> If I in one situation would use a one dimension array and in another
> situation a mutiple dimension array how would i do?
> I have tried with findgen(x), where x could be an integer or an array.
> But when x is an array IDL gives me following error message:
> IDL> r=fltarr(x)+10
> % FLTARR: Expression must be a scalar in this context: X.
> % Execution halted at: $MAIN$
>
> where x is:
> IDL> print,x
> 5 7
> IDL> help,x
> X INT = Array[2]
> Is it posible to make an mutiple dimensional array by using a variable
> such as findgen(x)?
> --
> Thanx in advance
> Martin Skou Andersen
> E-mail: Skou@fys.ku.dk
> Martin_Skou@mail.tele.dk
> Homepage: http://www.fys.ku.dk/~skou/
|
|
|
|
| Re: findgen([variable]) [message #22901 is a reply to message #22900] |
Tue, 12 December 2000 09:28 |
Craig Markwardt
Messages: 1869
Registered: November 1996
|
Senior Member |
|
|
Martin Skou Andersen <skou@fys.ku.dk> writes:
> Is it posible to make an mutiple dimensional array by using a variable
> such as findgen(x)?
This is virtually impossible, right? You are asking for each row of
the resulting array to be of a different length. That's not really a
2 dimensional array is it?
Craig
--
------------------------------------------------------------ --------------
Craig B. Markwardt, Ph.D. EMAIL: craigmnet@cow.physics.wisc.edu
Astrophysics, IDL, Finance, Derivatives | Remove "net" for better response
------------------------------------------------------------ --------------
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
