comp.lang.idl-pvwave archive: archive

Home » Public Forums » archive » clustering

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: Clustering [message #78244 is a reply to message #78177]

Tue, 01 November 2011 12:16

Jeremy Bailin
Messages: 618
Registered: April 2008

Senior Member

On 10/31/11 11:40 AM, Kai Muehlbauer wrote:
> Hi all,
>
> I took a big step forward.
>
> I slightly changed my histograms. I reduced the number of bins by
> increasing the binsize. I cut off noise before the histograms which also
> reduces number of bins. Then I fill the histograms in an array similar
> to Mario is doing.
>
> FOR K=0L, 1999 DO BEGIN
> FOR J=0L, 359 DO BEGIN
> array = REFORM(source[K,*,J)]
> hist_arr = HISTOGRAM(array,BINSIZE=0.5, MAX=7.5, MIN=0)
> Array[*,K*360L+J] = hist_arr
> ENDFOR
> ENDFOR
>
> Then the weights for 10 Clusters are calculated and CLUSTER is called
>
> weights = CLUST_WTS(array2, N_CLUSTERS = 10)
> tmp_result = CLUSTER(array1, weights, N_CLUSTERS = 10)
>
> Then the data needs REFORMing
>
> result1 = REFORM(tmp_result,360,2000)
>
> and in my case the dimensions need to be interchanged
>
> FOR I=0,range1 - 1 DO BEGIN
> result[I,*] = result1[*,I]
>
> Anyway the results were not useful. I noticed that a great deal (about
> 90 percent) of the histograms could be grouped into one cluster. So I
> reduced the histograms used to calculate the weights to a reasonable
> amount, to get better weights also for the remaining 10 percent.
>
> After that I get quite usable clusters of my data. I think with a little
> tweaking there should be even better results.
>
> There should also be a speedup possible in the above code. But I'am
> still in the learning phase, so a little help is appreciated. I still
> struggle with those dimensions.
>
> Thanks Mario for providing your example. I tried this but got an out of
> memory error while calculating the distance matrix. But that was before
> my reduction of histogram number of bins. I will test this later and
> come back with some results in november ;-)
>
> Cheers,
> Kai
>

You can probably use JD's HIST_ND to get rid of those for loops, which
should speed things up.

-Jeremy.

Report message to a moderator

[Message index]

		clustering By: nivedita.raghunath on Tue, 17 July 2007 08:36
		Re: clustering By: Vince Hradil on Thu, 19 July 2007 10:30
		Re: clustering By: nivedita.raghunath on Thu, 19 July 2007 09:12
		Re: clustering By: Vince Hradil on Thu, 19 July 2007 13:46
		Re: clustering By: Conor on Wed, 18 July 2007 12:07
		Re: clustering By: Vince Hradil on Mon, 23 July 2007 11:22
		Re: clustering By: James Kuyper on Mon, 23 July 2007 11:16
		Re: clustering By: little davey on Mon, 23 July 2007 10:14
		Re: clustering By: little davey on Mon, 23 July 2007 09:08
		Re: clustering By: Conor on Mon, 23 July 2007 06:49
		Re: clustering By: nivedita.raghunath on Mon, 23 July 2007 06:45
		Re: Clustering By: Kai Mühlbauer on Tue, 01 November 2011 03:01
		Re: Clustering By: Kai Muehlbauer on Mon, 31 October 2011 08:40
		Re: Clustering By: Jeremy Bailin on Tue, 01 November 2011 12:16
		Re: Clustering By: kidpix on Mon, 31 October 2011 04:56
		Re: Clustering By: Kai Mühlbauer on Fri, 28 October 2011 10:07
		Re: clustering By: kisCA on Wed, 07 March 2012 13:37
		Re: clustering By: Russell[1] on Wed, 07 March 2012 09:32

Previous Topic:	Re: Move data to structure fields without copying
Next Topic:	Annotate Uncertainty [Upper & Lower] on a plot?

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Thu Oct 09 22:29:01 PDT 2025

Total time taken to generate the page: 2.15981 seconds