| Re: concatenate arrays of different sizes [message #59427] |
Thu, 27 March 2008 09:42  |
R.G. Stockwell
Messages: 363
Registered: July 1999
|
Senior Member |
|
|
"vino" <astrocrazy@gmail.com> wrote in message
news:98fa711f-7d7e-4131-a70a-a75dfe22850d@d4g2000prg.googleg roups.com...
> Hello Everyone,
> I am trying to concatenate arrays of different sizes into a single
> variable.The problem is as follows:
> I am tracking about 10,000 stars through a period of six months. I
> have a single variable containing the intensity of all stars with
> zeros when the star is not in the field of view. I ran into memory
> problems when i try to track for longer periods.
What is time sampling on those?
You have intensity (float?) for 10k stars, by how many times in that siz
months?
Every ten minutes, fits into my winxp laptop just fine.
ie a = fltarr(10000,6*30*24*6)
Some suggestions:
Perhaps you could downsample the intensity time series - do
you really need that high time resultion?
You could categorize the stars into groups (based on quadrant in
the sky, or on magnitude) and analyze the groups seperately.
You make a pointer array for each star, 10k pointers where each pointer
points to a strucutre which holds the time and intensity
for when it is in view.
Cheers,
bob
|
|
|
|
|
|
| Re: concatenate arrays of different sizes [message #59511 is a reply to message #59427] |
Fri, 28 March 2008 02:01  |
vino
Messages: 36
Registered: March 2008
|
Member |
|
|
On Mar 27, 9:42 pm, "R.G. Stockwell" <notha...@noemail.com> wrote:
> "vino" <astrocr...@gmail.com> wrote in message
>
> news:98fa711f-7d7e-4131-a70a-a75dfe22850d@d4g2000prg.googleg roups.com...
>
>> Hello Everyone,
>> I am trying to concatenate arrays of different sizes into a single
>> variable.The problem is as follows:
>> I am tracking about 10,000 stars through a period of six months. I
>> have a single variable containing the intensity of all stars with
>> zeros when the star is not in the field of view. I ran into memory
>> problems when i try to track for longer periods.
>
> What is time sampling on those?
>
> You have intensity (float?) for 10k stars, by how many times in that siz
> months?
>
> Every ten minutes, fits into my winxp laptop just fine.
>
> ie a = fltarr(10000,6*30*24*6)
>
> Some suggestions:
> Perhaps you could downsample the intensity time series - do
> you really need that high time resultion?
>
> You could categorize the stars into groups (based on quadrant in
> the sky, or on magnitude) and analyze the groups seperately.
>
> You make a pointer array for each star, 10k pointers where each pointer
> points to a strucutre which holds the time and intensity
> for when it is in view.
>
> Cheers,
> bob
Hello Bob,
Thank you for your suggestion.
The cadence of this particular instrument is 40 minutes. And since i
am looking for transiting planets, this high cadence is very essential
to me.
The intensity array i am using contains flux for 3 different apertures
and hence the larger size.
As you suggested, catagorising it by quadrants was one of my ideas and
the other is to build a database. But now i will learn how to use
pointers and store it.
Thank you so much.
To Vince Hardi:
Thank you so much. Didnt think of it. Will try that.
regards,
vino
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
