Cluster Pattern Filter

Postby PadawanLotto » Sun Jun 01, 2008 11:51 am

[quote=Midas]
Teoman has made up some fancy errors.

Audit of your response error file shows complete disregard of the details of our original check file!.

[1] Our original file column for cluster pattern 3-1-1-1 shows (60 Items) with arbitary value predictor indexes (371 to 420) . All the patterns are there and complete.

Teoman's column has copied his name onto our original file (3-1-1-1 ) (120 Items) cluster pattern column giving the fake impression that there are mistakes in our data. He presents his name on our column and then gives us a deliberate error column with 50 items jumbled together?.

[2] Our original file column for cluster pattern 3-2-1 shows (120) items with arbitary value predictor indexes(251- 370). All the patterns are there and complete.

Teoman again has COPIED his name against our original column for the cluster pattern (3-2-1), then posted a deliberate fake column with patterns that don't exist from value predictor index (300) .He plants additional clusters of 3,4,5,6,7,8,9 for clusters which don't exist for this pattern. Worse still he wants you to believe we are so stupid that we have made these errors.

Why do you waste our time?.

Jerzy/Joe I would urge you to re-consider if you want to develop this filter further as I see it serves no purpose for users like Teoman.
[/quote]Midas, I don't know about cluster patterns but looking at your cluster pattern 3-1-1-1 and following the sequences going down the list at 310011 to 301101 was your first mistake, you have left out 301110 in your sequence of patterns. Teomans sequences look to be correct to me. In the cluster pattern 3-2-1 right after 120030 you have 120031 and from that point the rest of the sequence patterns seem to be wrong as the sequence patterns should equal 6. I downloaded IPSGRID-1_new.xls from your post not Teomans.
PadawanLotto
 
Posts: 1584
Joined: Sat Mar 25, 2006 10:33 pm
Location: Dayton, OH

Postby teoman » Sun Jun 01, 2008 12:31 pm

[quote=Midas]
Teoman has made up some fancy errors.

Audit of your response error file shows complete disregard of the details of our original check file!.

[1] Our original file column for cluster pattern 3-1-1-1 shows (60 Items) with arbitary value predictor indexes (371 to 420) . All the patterns are there and complete.

Teoman's column has copied his name onto our original file (3-1-1-1 ) (120 Items) cluster pattern column giving the fake impression that there are mistakes in our data. He presents his name on our column and then gives us a deliberate error column with 50 items jumbled together?.

[2] Our original file column for cluster pattern 3-2-1 shows (120) items with arbitary value predictor indexes(251- 370). All the patterns are there and complete.
Teoman again has COPIED his name against our original column for the cluster pattern (3-2-1), then posted a deliberate fake column with patterns that don't exist from value predictor index (300) .He plants additional clusters of 3,4,5,6,7,8,9 for clusters which don't exist for this pattern. Worse still he wants you to believe we are so stupid that we have made these errors.

Why do you waste our time?.

Jerzy/Joe I would urge you to re-consider if you want to develop this filter further as I see it serves no purpose for users like Teoman.
[/quote]

Midas !....Midas!......:vogel: :wand:
teoman
 
Posts: 191
Joined: Sun Apr 27, 2008 11:27 pm

Postby Midas » Sun Jun 01, 2008 12:39 pm

Good Point,

I just checked the file that Jerzy sent me against my master file patterns that I posted from him against Teoman's comments. My updated original excel file patterns are those which I based my commentary. Jerzy has appeared to have introduced some errors for my original cluster pattern (3-2-1 and 3-1-1-1 looks wrong), which quite rightly Teoman has highlighted as errors. My apologies.

In any case this irritating weekend has been extremely dissapointing, invariably not the right frame of mind to do any productive work. I am frankly very fed up with doing any further development on such filters.

Regards
Midas
Midas
 
Posts: 144
Joined: Wed Apr 19, 2006 7:20 pm

Postby teoman » Sun Jun 01, 2008 1:09 pm

[quote=Midas]
Good Point,

I just checked the file that Jerzy sent me against my master file patterns that I posted from him against Teoman's comments. My updated original excel file patterns are those which I based my commentary. Jerzy has appeared to have introduced some errors for my original cluster pattern (3-2-1 and 3-1-1-1 looks wrong), which quite rightly Teoman has highlighted as errors. My apologies.

In any case this irritating weekend has been extremely dissapointing, invariably not the right frame of mind to do any productive work. I am frankly very fed up with doing any further development on such filters.

Regards
Midas
[/quote]

Midas!.....Congratulations!!!!......
teoman
 
Posts: 191
Joined: Sun Apr 27, 2008 11:27 pm

Postby Jerzy » Sun Jun 01, 2008 3:35 pm

Midas,

I apologize for the errors in the IPSGRID-1_new.xls file and a disappointment they cause to you and other people. It is also very disappointing to me.

At your request I reviewed your Excel file and made some changes in it. Unfortunately, during the final editing (removing provisional notes) the file became corrupt, and unusual errors, eg pattern 120097, were introduced.

Before sending this file back to you, I looked for possible errors in it, but I have not noticed them, even though they were obvious.

Jerzy
Jerzy
 
Posts: 811
Joined: Mon Sep 18, 2006 5:26 am

Postby Jerzy » Sun Jun 01, 2008 3:39 pm

teoman,

Thank you for reporting the errors in the IPSGRID-1_new.xls file and correcting them.

Jerzy
Jerzy
 
Posts: 811
Joined: Mon Sep 18, 2006 5:26 am

Postby Midas » Sun Jun 01, 2008 3:44 pm

No problem and thanks for the reply Jerzy,

I appreciate your honesty and must admit I should have looked more carefully at the file before posting. All the rest of the many cluster patterns line up perfectly. In any case without any statistics we are stuck. !.

Best Regards
Midas
Midas
 
Posts: 144
Joined: Wed Apr 19, 2006 7:20 pm

Postby CARBOB » Sun Jun 01, 2008 3:46 pm

[quote=Midas]
No problem and thanks for the reply Jerzy,

I appreciate your honesty and must admit I should have looked more carefully at the file before posting. All the rest of the many cluster patterns line up perfectly. In any case without any statistics we are stuck. !.

Best Regards
Midas
[/quote]


Everyone needs to remember the most important thing!! The error was caught and corrected, no harm done!! Would someone please post the corrected Excel worksheet??
CARBOB
 
Posts: 874
Joined: Thu Jun 29, 2006 10:34 pm

Postby Jerzy » Sun Jun 01, 2008 4:02 pm

Carbob,

I will correct the errors and check everything again, and then I will post it on the Forum, probably in 2 days, if Midas agrees.

Jerzy
Jerzy
 
Posts: 811
Joined: Mon Sep 18, 2006 5:26 am

Postby Midas » Sun Jun 01, 2008 10:18 pm

Thanks Jerzy,

This frees up time for me to run the algorithm soffware through my party piece. I hope to have it singing tunes shortly. Appreciate all your hard efforts!.

I am very keen on your feedback on my work and of course would also be delighted to receive commentary from CARBOB, Teoman and Padawan.
If anybody else wishes to add to the deliverables that would be great.

Best Regards
Midas
Midas
 
Posts: 144
Joined: Wed Apr 19, 2006 7:20 pm

Postby Jerzy » Tue Jun 03, 2008 1:12 pm

Midas,

I corrected the errors in the Excel file. And I believe all errors are fixed and that I can post this file now.

The file shows all possible cluster patterns and configurations of clusters and isolated numbers. They are arranged in the tables in descending order.

The Arbitrary Value Predictor Identifier is a kind of identification number for the cluster or the combination of clusters and isolated numbers (arbitrary, because any number could be assigned).

Zeros in the cluster patterns reflect your idea to put each cluster or isolated number into a separate bin. 6 bins are needed for this. A zero in the cluster pattern means that a bin is empty.

Best Regards

Jerzy
You do not have the required permissions to view the files attached to this post.
Jerzy
 
Posts: 811
Joined: Mon Sep 18, 2006 5:26 am

Postby CARBOB » Tue Jun 03, 2008 3:04 pm

Jerzy, thanks for re-posting!!
CARBOB
 
Posts: 874
Joined: Thu Jun 29, 2006 10:34 pm

Postby Midas » Thu Jun 05, 2008 8:45 pm

Thanks Jerzy,

Stan/Joe we have achieved quite a lot of work with this new spatial filter, the cluster filter very much needs statistics to function properly. Can you help?. :listen:
Midas
 
Posts: 144
Joined: Wed Apr 19, 2006 7:20 pm

Postby Joe » Mon Jun 09, 2008 3:56 pm

yeah midas, I feel little lazy about this filter, i don't know why i am taking too much time for this, :wand: , we can have this filter asap, thanks,
Joe
 
Posts: 1227
Joined: Mon Sep 18, 2006 6:51 pm
Location: Petaling Jaya, Malaysia

Cluster Pattern Analyser & Filter

Postby Midas » Thu Aug 06, 2009 3:00 pm

Hello Joe and fellow EL users,

Just a quick note to ask if you have had further thoughts on the cluster filter?.

The temperature here has almost reached an all time high, good time to jump into the pool and cool off slowly...!. ;-)

Regards
Midas
Midas
 
Posts: 144
Joined: Wed Apr 19, 2006 7:20 pm

Previous

Return to Plugin Development

Who is online

Users browsing this forum: No registered users and 5 guests

cron