R help - May 2006 - How-Understand-Kmeans-Cluster!!

Hi All,

I am using RGui 2.3.0 library(stats) and using K-means Clustering to cluster
data.matrix.(DDset)

DDset (matrix 575 x 2)

Exams (character) Slot (numeric)
 AAW326                35
 BWA345                45
 ...............             .....

  ZXE 453               67

DDset[,2]

 [1] 29 37 24  7 22 22 21 24 26 31 39 21 17  7 24 37 26 33  7 31 39 29 24
20  2
 [26] 40  8 16  1 31  6 32  2 13 16 16 40 16 40 16 33  9 20  2 22 36 40  3
25  2
 [51] 22 40 32  2  2 13 21 40  2 22 18 13 26 32  8 30 39 28 28 35 14 14  5
8 18
 [76] 20 38 26 40 22 18 28 24  5 26 21 17 14 28  1 23 21 37 22 31 13  6  5
24 39
[101] 34  5 32 28 32 34 22  7 38 38  6 14 21 28  7 36 33 21 28 38 36  5  7
38  2
[126] 40 38  5  2  7 38 33  2  7 26 40 38 16 18 14 38 13  6 40 38 14 14 33
1 33
[151] 36 31 36  3 30 16  6 23 37 25 26 22 24 28 32 23 30 15  3  8  3 26 39
13  8
[176]  5  5 32  6  3 28 34 30  5 26 28 13  2  5 17 23 18 33 38 22 40  6 32
29 26
[201] 36  5  1  3 37 36 16 24 30 26  7 29  2  8 20 13 22 28 15 32 23 34 27
13 39
[226] 32  2 16  1 37 24  2 39 28 20 18 30 26 19 14 21 32 28 24  6 20 36 21
32 27
[251]  6 29 35 37 17 23 21 15 38 15 34 17 32 22 13  7 23 13 34 39  5 25 21
14 29
[276] 40 38  6 21  5 33 33 14  1 23 21 23 33 32  4 26 36 17  6 20 29  6 26
36 34
[301] 20 39  2 36  1 20 28 36 30 32 23 23  1 24 22 28  8  8 36 35 26 13 30
13 36
[326] 37 17  1 30 18 13  9  4  6 39 38 15 32 18 30 30 13  1 35  3 30 15 21
29  8
[351] 25 29  1 22 32 31 33 24 34 39 35 23 31 33 26 30 26 25 40 33  9 38  9
20 34
[376] 34  5 30 33  9 38  9  9 20 34 34 17  8  3 37  8 29 13 16 18 22 39 16
22 16
[401] 24 16 30 27 37 30  9 40 28 24 39 36  5  7  7 32 16 30 35 26 22 23 13
34 25
[426] 21 37 21 34  6 32  6  8 36  1 21 40 13 24 18 33 38  4 25  3 34 17 37
13 34
[451] 39 27 38 25 36 22 26 16 25 21 36  5 30 39 18 14 23  2  6  6 32 28 40
30 20
[476] 22  1 34  1 27 24 31 24 40 20  5  3 26 27 38  6 13 20 27 29 17 25 29
22 33
[501]  9  9 31 28 18 16 35 40  4 13 36 37 21 26 32 16 25 34 21 27 32  1  5
6 36
[526]  1 24 38 13 16 34 30 24 24  7 21  6  7 33  6  2 26 38 32  3  5 28 26
28 38
[551] 27 22 26 29 27  7  5 13 23 26 21  3 38  5  8 28  8 22  5 33 24 18 21
32 13
>
I already have pre-process of the data i.e *Exam AAW326* is assigned to the
*slot-29 *which is the dataset
DDset. By using Kmeans-Clustering i am grouping the dataset accordingly. How
can i do the clustering
from the datset (DDset)? How can i understand the *similarities and
dis-similarities*? Will Kmeans-Clustering
helps on this problem? Where can i get *more tutorials* on
kmeans-clustering?

Thanks for reading this email.

Thanks in Advance

JJ
---












-- 
Lecturer J. Joshua Thomas
KDU College Penang Campus
Research Student,
University Sains Malaysia

	[[alternative HTML version deleted]]

Hi all,

Can someone help me with the following expectation in a closed form?

E(1/(X+c))

where X ~ Chi-square(n) and c is a constant.

Thanks.

	[[alternative HTML version deleted]]

Joshua,

If you want a comparison of different distance measures try
> the paper of *Johann* *Bacher* who is a specialists in *cluster* methods. 
> He published an in-depth text on his web site 
> http://www.soziologie.wiso.uni-erlangen.de/koeln/
> 
> Regards,
F. Thomas


j.joshua thomas wrote:
> Hi All,
>
> I am using RGui 2.3.0 library(stats) and using K-means Clustering to
cluster
> data.matrix.(DDset)
>
> DDset (matrix 575 x 2)
>
> Exams (character) Slot (numeric)
>  AAW326                35
>  BWA345                45
>  ...............             .....
>
>   ZXE 453               67
>
> DDset[,2]
>
>  [1] 29 37 24  7 22 22 21 24 26 31 39 21 17  7 24 37 26 33  7 31 39 29 24
> 20  2
>  [26] 40  8 16  1 31  6 32  2 13 16 16 40 16 40 16 33  9 20  2 22 36 40  3
> 25  2
>  [51] 22 40 32  2  2 13 21 40  2 22 18 13 26 32  8 30 39 28 28 35 14 14  5
> 8 18
>  [76] 20 38 26 40 22 18 28 24  5 26 21 17 14 28  1 23 21 37 22 31 13  6  5
> 24 39
> [101] 34  5 32 28 32 34 22  7 38 38  6 14 21 28  7 36 33 21 28 38 36  5  7
> 38  2
> [126] 40 38  5  2  7 38 33  2  7 26 40 38 16 18 14 38 13  6 40 38 14 14 33
> 1 33
> [151] 36 31 36  3 30 16  6 23 37 25 26 22 24 28 32 23 30 15  3  8  3 26 39
> 13  8
> [176]  5  5 32  6  3 28 34 30  5 26 28 13  2  5 17 23 18 33 38 22 40  6 32
> 29 26
> [201] 36  5  1  3 37 36 16 24 30 26  7 29  2  8 20 13 22 28 15 32 23 34 27
> 13 39
> [226] 32  2 16  1 37 24  2 39 28 20 18 30 26 19 14 21 32 28 24  6 20 36 21
> 32 27
> [251]  6 29 35 37 17 23 21 15 38 15 34 17 32 22 13  7 23 13 34 39  5 25 21
> 14 29
> [276] 40 38  6 21  5 33 33 14  1 23 21 23 33 32  4 26 36 17  6 20 29  6 26
> 36 34
> [301] 20 39  2 36  1 20 28 36 30 32 23 23  1 24 22 28  8  8 36 35 26 13 30
> 13 36
> [326] 37 17  1 30 18 13  9  4  6 39 38 15 32 18 30 30 13  1 35  3 30 15 21
> 29  8
> [351] 25 29  1 22 32 31 33 24 34 39 35 23 31 33 26 30 26 25 40 33  9 38  9
> 20 34
> [376] 34  5 30 33  9 38  9  9 20 34 34 17  8  3 37  8 29 13 16 18 22 39 16
> 22 16
> [401] 24 16 30 27 37 30  9 40 28 24 39 36  5  7  7 32 16 30 35 26 22 23 13
> 34 25
> [426] 21 37 21 34  6 32  6  8 36  1 21 40 13 24 18 33 38  4 25  3 34 17 37
> 13 34
> [451] 39 27 38 25 36 22 26 16 25 21 36  5 30 39 18 14 23  2  6  6 32 28 40
> 30 20
> [476] 22  1 34  1 27 24 31 24 40 20  5  3 26 27 38  6 13 20 27 29 17 25 29
> 22 33
> [501]  9  9 31 28 18 16 35 40  4 13 36 37 21 26 32 16 25 34 21 27 32  1  5
> 6 36
> [526]  1 24 38 13 16 34 30 24 24  7 21  6  7 33  6  2 26 38 32  3  5 28 26
> 28 38
> [551] 27 22 26 29 27  7  5 13 23 26 21  3 38  5  8 28  8 22  5 33 24 18 21
> 32 13
>   
>
> I already have pre-process of the data i.e *Exam AAW326* is assigned to the
> *slot-29 *which is the dataset
> DDset. By using Kmeans-Clustering i am grouping the dataset accordingly.
How
> can i do the clustering
> from the datset (DDset)? How can i understand the *similarities and
> dis-similarities*? Will Kmeans-Clustering
> helps on this problem? Where can i get *more tutorials* on
> kmeans-clustering?
>
> Thanks for reading this email.
>
> Thanks in Advance
>
> JJ
> ---
>
>
>
>
>
>
>
>
>
>
>
>
>   

-- 
..........................................
Dr. Frank Thomas
FTR Internet Research
93110 Rosny-sous-Bois
France



--