i want to study Singular Value Decomposition algorithms;
I also have a book called mahout in action,but i can`t found sth about this algorithm;
is there someplace introduce how to use the method?
till now DistributedLanczosSolver  is not a mapreduce method
org.apache.mahout.math.hadoop.decomposer.DistributedLanczosSolver = svd

http://ocw.mit.edu/courses/mathematics/18-06-linear-algebra-spring-2010/video-lectures/

On 29.08.2011 09:15, myn wrote:
> i want to study Singular Value Decomposition algorithms;
> I also have a book called mahout in action,but i can`t found sth about this algorithm;
> is there someplace introduce how to use the method?
> till now DistributedLanczosSolver  is not a mapreduce method
> org.apache.mahout.math.hadoop.decomposer.DistributedLanczosSolver = svd

Command line arguments are found here:
https://cwiki.apache.org/MAHOUT/dimensional-reduction.html
I wrote a quick tutorial on how to prepare sparse matrices as input to
Mahout SVD here:
http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html

Let me know if you have further questions.

2011/8/29 myn <[EMAIL PROTECTED]>

> i want to study Singular Value Decomposition algorithms;
> I also have a book called mahout in action,but i can`t found sth about this
> algorithm;
> is there someplace introduce how to use the method?
> till now DistributedLanczosSolver  is not a mapreduce method
> org.apache.mahout.math.hadoop.decomposer.DistributedLanczosSolver = svd

thanks
But could you send the content ofhttp://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html to me ?
I can`t open it  in china .
在 2011-08-29 15:29:46，"Sebastian Schelter" <[EMAIL PROTECTED]> 写道：
>http://ocw.mit.edu/courses/mathematics/18-06-linear-algebra-spring-2010/video-lectures/
>
>On 29.08.2011 09:15, myn wrote:
>> i want to study Singular Value Decomposition algorithms;
>> I also have a book called mahout in action,but i can`t found sth about this algorithm;
>> is there someplace introduce how to use the method?
>> till now DistributedLanczosSolver  is not a mapreduce method
>> org.apache.mahout.math.hadoop.decomposer.DistributedLanczosSolver = svd
>

thanks
But could you send the content ofhttp://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html to me ?
I can`t open it  in china .

At 2011-08-29 15:29:40,"Danny Bickson" <[EMAIL PROTECTED]> wrote:
>Command line arguments are found here:
>https://cwiki.apache.org/MAHOUT/dimensional-reduction.html
>I wrote a quick tutorial on how to prepare sparse matrices as input to
>Mahout SVD here:
>http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html
>
>Let me know if you have further questions.
>
>2011/8/29 myn <[EMAIL PROTECTED]>
>
>> i want to study Singular Value Decomposition algorithms;
>> I also have a book called mahout in action,but i can`t found sth about this
>> algorithm;
>> is there someplace introduce how to use the method?
>> till now DistributedLanczosSolver  is not a mapreduce method
>> org.apache.mahout.math.hadoop.decomposer.DistributedLanczosSolver = svd

2011/8/29 myn <[EMAIL PROTECTED]>:
> thanks
> But could you send the content ofhttp://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html to me ?

(You asked the same thing twice with only 6 minutes between)
Try this:

http://translate.google.com/translate?js=n&prev=_t&hl=en&ie=UTF-8&layout=2&eotf=1&sl=zh-CN&tl=en&u=http%3A%2F%2Fbickson.blogspot.com%2F2011%2F02%2Fmahout-svd-matrix-factorization.html

Or look for similar services that take a URL and then repeat its contents...

For SVD, see also
https://cwiki.apache.org/MAHOUT/dimensional-reduction.html and
https://cwiki.apache.org/MAHOUT/svd-singular-value-decomposition.html
... the later has links to some mail threads here too:
http://mail-archives.apache.org/mod_mbox/mahout-user/201102.mbox/%[EMAIL PROTECTED]%3E

Dan

ps. for general learning materials, look beyond Mahout. Eg. this in
Ruby might help:
http://www.igvita.com/2007/01/15/svd-recommendation-system-in-ruby/

Mahout - SVD matrix factorization - formatting input matrix
 Converting Input Format into Mahout's SVD Distributed Matrix Factorization
Solver

Purpose
The code below, converts a matrix from csv format:
<from row>,<to col>,<value>\n
Into Mahout's SVD solver format.
For example,
The 3x3 matrix:
0    1.0 2.1
3.0  4.0 5.0
-5.0 6.2 0
Will be given as input in a csv file as:
1,0,3.0
2,0,-5.0
0,1,1.0
1,1,4.0
2,1,6.2
0,2,2.1
1,2,5.0

NOTE: I ASSUME THE MATRIX IS SORTED BY THE COLUMNS ORDER
This code is based on code by Danny Leshem, ContextIn.

Command line arguments:
 args[0] - path to csv input file
args[1] - cardinality of the matrix (number of columns)
args[2] - path the resulting Mahout's SVD input file

Method:
The code below, goes over the csv file, and for each matrix column, creates
a SequentialAccessSparseVector which contains all the non-zero row entries
for this column.
Then it appends the column vector to file.

Compilation:
Copy the java code below into an java file named Convert2SVD.java
Add to your IDE project path both Mahout and Hadoop jars. Alternatively, a
command line option for compilation is given below.
view plain<http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
print<http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
?<http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>

   1. import java.io.BufferedReader;
   2. import java.io.FileReader;
   3. import java.util.StringTokenizer;
   4.
   5. import org.apache.mahout.math.SequentialAccessSparseVector;
   6. import org.apache.mahout.math.Vector;
   7. import org.apache.mahout.math.VectorWritable;
   8. import org.apache.hadoop.conf.Configuration;
   9. import org.apache.hadoop.fs.FileSystem;
   10. import org.apache.hadoop.fs.Path;
   11. import org.apache.hadoop.io.IntWritable;
   12. import org.apache.hadoop.io.SequenceFile;
   13. import org.apache.hadoop.io.SequenceFile.CompressionType;
   14.
   15. /**
   16.  * Code for converting CSV format to Mahout's SVD format
   17.  * @author Danny Bickson, CMU
   18.
    * Note: I ASSUME THE CSV FILE IS SORTED BY THE COLUMN (NAMELY THE
SECOND FIELD).

   19.  *
   20.  */
   21.
   22. public class Convert2SVD {
   23.
   24.
   25.         public static int Cardinality;
   26.
   27.         /**
   28.          *
   29.          * @param args[0] - input csv file
   30.          * @param args[1] - cardinality (length of vector)
   31.          * @param args[2] - output file for svd
   32.          */
   33.         public static void main(String[] args){
   34.
   35. try {
   36.         Cardinality = Integer.parseInt(args[1]);
   37.         final Configuration conf = new Configuration();
   38.         final FileSystem fs = FileSystem.get(conf);
   39.         final
    SequenceFile.Writer writer = SequenceFile.createWriter(fs, conf, new
    Path(args[2]), IntWritable.class, VectorWritable.class
   , CompressionType.BLOCK);
   40.
   41.           final IntWritable key = new IntWritable();
   42.           final VectorWritable value = new VectorWritable();
   43.
   44.
   45.            String thisLine;
   46.
   47.            BufferedReader br = new BufferedReader(new
    FileReader(args[0]));
   48.            Vector vector = null;
   49.            int from = -1,to  =-1;
   50.            int last_to = -1;
   51.            float val = 0;
   52.            int total = 0;
   53.            int nnz = 0;
   54.            int e = 0;
   55.            int max_to =0;
   56.            int max_from = 0;
   57.
   58.            while ((thisLine = br.readLine()) != null) {
   // while loop begins here
   59.
   60.                  StringTokenizer st = new StringTokenizer(thisLine,
   ",");
   61.                  while(st.hasMoreTokens()) {
   62.                      from = Integer.parseInt(st.nextToken())-1;
   //convert from 1 based to zero based
   63.                      to = Integer.parseInt(st.nextToken())-1;
   //convert from 1 based to zero basd
   64.                      val = Float.parseFloat(st.nextToken());
   65.                      if (max_from < from) max_from = from;
   66.                      if (max_to < to) max_to = to;
   67.                      if (from < 0 || to < 0
    || to > Cardinality || val == 0.0)
   68.                          throw new NumberFormatException("wrong data"
    + from + " to: " + to + " val: " + val);
   69.                  }
   70.
   71.
   //we are working on an existing column, set non-zero rows in it
   72.                  if (last_to != to && last_to != -1){
   73.                      value.set(vector);
   74.
   75.                      writer.append(key, value);
   //write the older vector
   76.                      e+= vector.getNumNondefaultElements();
   77.                  }
   78.                  //a new column is observed, open a new vector for it

   79.                  if (last_to != to){
   80.                      vector = new
    SequentialAccessSparseVector(Cardinality);
   81.                      key.set(to); // open a new vector
   82.                      total++;
   83.                  }
   84.
   85.                  vector.set(from, val);
   86.                  nnz++;
   87.
   88.                  if (nnz % 1000000 == 0){
   89.                    System.out.println("Col" + total + " nnz: "
    + nnz);
   90.                  }
   91.                  last_to = to;
   92.
   93.           } // end while
   94.
   95.            value.set(vector);
   96.            writer.append(key,value);//write last row
   97.            e+= vector.getNumNondefaultElements();
   98.            total++;
   99.
   100.            writer.close();
   101.            System.out.println("Wrote a total of " + total + " cols "
    + " nnz: " + nnz);
   102.            if (e != nnz)
   103.                 System.err.println("Bug:missing edges! we only got"
    + e);
   104.
   105.            System.out.println("Highest column: " +

'R' also has an svd implementation, directly in the base package.

There are a few answers to your question:
1) What is SVD? The video lecture above will help. Also, searching for
'singular value decomposition' on Baidu finds a lot of basic explanations.
2) Why do you want it? It creates in on pass a few different unique
explanations of what is going on inside your dataset.
3) Mahout Distributed Matrix code, DistributedLanczos etc. are
implementations specifically for large-scale problems. There are sub-parts
of SVD that you may not need for your problem, and these jobs avoid some of
the work.

Until you have a solid grasp of what SVD can tell you, there is no point
trying the distributed mahout jobs. The SingularValueDecomposition class in
Mahout has served me well in my researches.

Lance

On Mon, Aug 29, 2011 at 12:50 AM, Danny Bickson <[EMAIL PROTECTED]>wrote:

>  Mahout - SVD matrix factorization - formatting input matrix
>  Converting Input Format into Mahout's SVD Distributed Matrix Factorization
> Solver
>
> Purpose
> The code below, converts a matrix from csv format:
> <from row>,<to col>,<value>\n
> Into Mahout's SVD solver format.
>
>
> For example,
> The 3x3 matrix:
> 0    1.0 2.1
> 3.0  4.0 5.0
> -5.0 6.2 0
>
>
> Will be given as input in a csv file as:
> 1,0,3.0
> 2,0,-5.0
> 0,1,1.0
> 1,1,4.0
> 2,1,6.2
> 0,2,2.1
> 1,2,5.0
>
> NOTE: I ASSUME THE MATRIX IS SORTED BY THE COLUMNS ORDER
> This code is based on code by Danny Leshem, ContextIn.
>
> Command line arguments:
>  args[0] - path to csv input file
> args[1] - cardinality of the matrix (number of columns)
> args[2] - path the resulting Mahout's SVD input file
>
> Method:
> The code below, goes over the csv file, and for each matrix column, creates
> a SequentialAccessSparseVector which contains all the non-zero row entries
> for this column.
> Then it appends the column vector to file.
>
> Compilation:
> Copy the java code below into an java file named Convert2SVD.java
> Add to your IDE project path both Mahout and Hadoop jars. Alternatively, a
> command line option for compilation is given below.
>
>
> view plain<
> http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
> print<
> http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
> ?<
> http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
>
>   1. import java.io.BufferedReader;
>   2. import java.io.FileReader;
>   3. import java.util.StringTokenizer;
>   4.
>   5. import org.apache.mahout.math.SequentialAccessSparseVector;
>   6. import org.apache.mahout.math.Vector;
>   7. import org.apache.mahout.math.VectorWritable;
>   8. import org.apache.hadoop.conf.Configuration;
>   9. import org.apache.hadoop.fs.FileSystem;
>   10. import org.apache.hadoop.fs.Path;
>   11. import org.apache.hadoop.io.IntWritable;
>   12. import org.apache.hadoop.io.SequenceFile;
>   13. import org.apache.hadoop.io.SequenceFile.CompressionType;
>   14.
>   15. /**
>   16.  * Code for converting CSV format to Mahout's SVD format
>   17.  * @author Danny Bickson, CMU
>   18.
>    * Note: I ASSUME THE CSV FILE IS SORTED BY THE COLUMN (NAMELY THE
> SECOND FIELD).
>
>   19.  *
>   20.  */
>   21.
>   22. public class Convert2SVD {
>   23.
>   24.
>   25.         public static int Cardinality;
>   26.
>   27.         /**
>   28.          *
>   29.          * @param args[0] - input csv file
>   30.          * @param args[1] - cardinality (length of vector)
>   31.          * @param args[2] - output file for svd
>   32.          */
>   33.         public static void main(String[] args){
>   34.
>   35. try {
>   36.         Cardinality = Integer.parseInt(args[1]);
>   37.         final Configuration conf = new Configuration();
>   38.         final FileSystem fs = FileSystem.get(conf);
>   39.         final
>    SequenceFile.Writer writer = SequenceFile.createWriter(fs, conf, new
>    Path(args[2]), IntWritable.class, VectorWritable.class
>   , CompressionType.BLOCK);

Lance Norskog
[EMAIL PROTECTED]

thanks a lot ,that is a quit good example for my study.
At 2011-08-29 15:50:36,"Danny Bickson" <[EMAIL PROTECTED]> wrote:
> Mahout - SVD matrix factorization - formatting input matrix
> Converting Input Format into Mahout's SVD Distributed Matrix Factorization
>Solver
>
>Purpose
>The code below, converts a matrix from csv format:
><from row>,<to col>,<value>\n
>Into Mahout's SVD solver format.
>
>
>For example,
>The 3x3 matrix:
>0    1.0 2.1
>3.0  4.0 5.0
>-5.0 6.2 0
>
>
>Will be given as input in a csv file as:
>1,0,3.0
>2,0,-5.0
>0,1,1.0
>1,1,4.0
>2,1,6.2
>0,2,2.1
>1,2,5.0
>
>NOTE: I ASSUME THE MATRIX IS SORTED BY THE COLUMNS ORDER
>This code is based on code by Danny Leshem, ContextIn.
>
>Command line arguments:
> args[0] - path to csv input file
>args[1] - cardinality of the matrix (number of columns)
>args[2] - path the resulting Mahout's SVD input file
>
>Method:
>The code below, goes over the csv file, and for each matrix column, creates
>a SequentialAccessSparseVector which contains all the non-zero row entries
>for this column.
>Then it appends the column vector to file.
>
>Compilation:
>Copy the java code below into an java file named Convert2SVD.java
>Add to your IDE project path both Mahout and Hadoop jars. Alternatively, a
>command line option for compilation is given below.
>
>
>view plain<http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
>print<http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
>?<http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
>
>   1. import java.io.BufferedReader;
>   2. import java.io.FileReader;
>   3. import java.util.StringTokenizer;
>   4.
>   5. import org.apache.mahout.math.SequentialAccessSparseVector;
>   6. import org.apache.mahout.math.Vector;
>   7. import org.apache.mahout.math.VectorWritable;
>   8. import org.apache.hadoop.conf.Configuration;
>   9. import org.apache.hadoop.fs.FileSystem;
>   10. import org.apache.hadoop.fs.Path;
>   11. import org.apache.hadoop.io.IntWritable;
>   12. import org.apache.hadoop.io.SequenceFile;
>   13. import org.apache.hadoop.io.SequenceFile.CompressionType;
>   14.
>   15. /**
>   16.  * Code for converting CSV format to Mahout's SVD format
>   17.  * @author Danny Bickson, CMU
>   18.
>    * Note: I ASSUME THE CSV FILE IS SORTED BY THE COLUMN (NAMELY THE
>SECOND FIELD).
>
>   19.  *
>   20.  */
>   21.
>   22. public class Convert2SVD {
>   23.
>   24.
>   25.         public static int Cardinality;
>   26.
>   27.         /**
>   28.          *
>   29.          * @param args[0] - input csv file
>   30.          * @param args[1] - cardinality (length of vector)
>   31.          * @param args[2] - output file for svd
>   32.          */
>   33.         public static void main(String[] args){
>   34.
>   35. try {
>   36.         Cardinality = Integer.parseInt(args[1]);
>   37.         final Configuration conf = new Configuration();
>   38.         final FileSystem fs = FileSystem.get(conf);
>   39.         final
>    SequenceFile.Writer writer = SequenceFile.createWriter(fs, conf, new
>    Path(args[2]), IntWritable.class, VectorWritable.class
>   , CompressionType.BLOCK);
>   40.
>   41.           final IntWritable key = new IntWritable();
>   42.           final VectorWritable value = new VectorWritable();
>   43.
>   44.
>   45.            String thisLine;
>   46.
>   47.            BufferedReader br = new BufferedReader(new
>    FileReader(args[0]));
>   48.            Vector vector = null;
>   49.            int from = -1,to  =-1;
>   50.            int last_to = -1;
>   51.            float val = 0;
>   52.            int total = 0;
>   53.            int nnz = 0;
>   54.            int e = 0;
>   55.            int max_to =0;
>   56.            int max_from = 0;
>   57.
>   58.            while ((thisLine = br.readLine()) != null) {
>   // while loop begins here
>   59.
>   60.                  StringTokenizer st = new StringTokenizer(thisLine,
>   ",");

the best way is to read the sorce code ;
 
@_@
At 2011-08-29 16:02:57,"Lance Norskog" <[EMAIL PROTECTED]> wrote:
>'R' also has an svd implementation, directly in the base package.
>
>There are a few answers to your question:
>1) What is SVD? The video lecture above will help. Also, searching for
>'singular value decomposition' on Baidu finds a lot of basic explanations.
>2) Why do you want it? It creates in on pass a few different unique
>explanations of what is going on inside your dataset.
>3) Mahout Distributed Matrix code, DistributedLanczos etc. are
>implementations specifically for large-scale problems. There are sub-parts
>of SVD that you may not need for your problem, and these jobs avoid some of
>the work.
>
>Until you have a solid grasp of what SVD can tell you, there is no point
>trying the distributed mahout jobs. The SingularValueDecomposition class in
>Mahout has served me well in my researches.
>
>Lance
>
>On Mon, Aug 29, 2011 at 12:50 AM, Danny Bickson <[EMAIL PROTECTED]>wrote:
>
>>  Mahout - SVD matrix factorization - formatting input matrix
>>  Converting Input Format into Mahout's SVD Distributed Matrix Factorization
>> Solver
>>
>> Purpose
>> The code below, converts a matrix from csv format:
>> <from row>,<to col>,<value>\n
>> Into Mahout's SVD solver format.
>>
>>
>> For example,
>> The 3x3 matrix:
>> 0    1.0 2.1
>> 3.0  4.0 5.0
>> -5.0 6.2 0
>>
>>
>> Will be given as input in a csv file as:
>> 1,0,3.0
>> 2,0,-5.0
>> 0,1,1.0
>> 1,1,4.0
>> 2,1,6.2
>> 0,2,2.1
>> 1,2,5.0
>>
>> NOTE: I ASSUME THE MATRIX IS SORTED BY THE COLUMNS ORDER
>> This code is based on code by Danny Leshem, ContextIn.
>>
>> Command line arguments:
>>  args[0] - path to csv input file
>> args[1] - cardinality of the matrix (number of columns)
>> args[2] - path the resulting Mahout's SVD input file
>>
>> Method:
>> The code below, goes over the csv file, and for each matrix column, creates
>> a SequentialAccessSparseVector which contains all the non-zero row entries
>> for this column.
>> Then it appends the column vector to file.
>>
>> Compilation:
>> Copy the java code below into an java file named Convert2SVD.java
>> Add to your IDE project path both Mahout and Hadoop jars. Alternatively, a
>> command line option for compilation is given below.
>>
>>
>> view plain<
>> http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
>> print<
>> http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
>> ?<
>> http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
>>
>>   1. import java.io.BufferedReader;
>>   2. import java.io.FileReader;
>>   3. import java.util.StringTokenizer;
>>   4.
>>   5. import org.apache.mahout.math.SequentialAccessSparseVector;
>>   6. import org.apache.mahout.math.Vector;
>>   7. import org.apache.mahout.math.VectorWritable;
>>   8. import org.apache.hadoop.conf.Configuration;
>>   9. import org.apache.hadoop.fs.FileSystem;
>>   10. import org.apache.hadoop.fs.Path;
>>   11. import org.apache.hadoop.io.IntWritable;
>>   12. import org.apache.hadoop.io.SequenceFile;
>>   13. import org.apache.hadoop.io.SequenceFile.CompressionType;
>>   14.
>>   15. /**
>>   16.  * Code for converting CSV format to Mahout's SVD format
>>   17.  * @author Danny Bickson, CMU
>>   18.
>>    * Note: I ASSUME THE CSV FILE IS SORTED BY THE COLUMN (NAMELY THE
>> SECOND FIELD).
>>
>>   19.  *
>>   20.  */
>>   21.
>>   22. public class Convert2SVD {
>>   23.
>>   24.
>>   25.         public static int Cardinality;
>>   26.
>>   27.         /**
>>   28.          *
>>   29.          * @param args[0] - input csv file
>>   30.          * @param args[1] - cardinality (length of vector)
>>   31.          * @param args[2] - output file for svd
>>   32.          */
>>   33.         public static void main(String[] args){
>>   34.
>>   35. try {
>>   36.         Cardinality = Integer.parseInt(args[1]);
>>   37.         final Configuration conf = new Configuration();

thanks evey body , my chinese english ,
At 2011-08-29 19:03:59,myn <[EMAIL PROTECTED]> wrote:

the best way is to read the sorce code ;
 
@_@
At 2011-08-29 16:02:57,"Lance Norskog" <[EMAIL PROTECTED]> wrote:
>'R' also has an svd implementation, directly in the base package.
>
>There are a few answers to your question:
>1) What is SVD? The video lecture above will help. Also, searching for
>'singular value decomposition' on Baidu finds a lot of basic explanations.
>2) Why do you want it? It creates in on pass a few different unique
>explanations of what is going on inside your dataset.
>3) Mahout Distributed Matrix code, DistributedLanczos etc. are
>implementations specifically for large-scale problems. There are sub-parts
>of SVD that you may not need for your problem, and these jobs avoid some of
>the work.
>
>Until you have a solid grasp of what SVD can tell you, there is no point
>trying the distributed mahout jobs. The SingularValueDecomposition class in
>Mahout has served me well in my researches.
>
>Lance
>
>On Mon, Aug 29, 2011 at 12:50 AM, Danny Bickson <[EMAIL PROTECTED]>wrote:
>
>>  Mahout - SVD matrix factorization - formatting input matrix
>>  Converting Input Format into Mahout's SVD Distributed Matrix Factorization
>> Solver
>>
>> Purpose
>> The code below, converts a matrix from csv format:
>> <from row>,<to col>,<value>\n
>> Into Mahout's SVD solver format.
>>
>>
>> For example,
>> The 3x3 matrix:
>> 0    1.0 2.1
>> 3.0  4.0 5.0
>> -5.0 6.2 0
>>
>>
>> Will be given as input in a csv file as:
>> 1,0,3.0
>> 2,0,-5.0
>> 0,1,1.0
>> 1,1,4.0
>> 2,1,6.2
>> 0,2,2.1
>> 1,2,5.0
>>
>> NOTE: I ASSUME THE MATRIX IS SORTED BY THE COLUMNS ORDER
>> This code is based on code by Danny Leshem, ContextIn.
>>
>> Command line arguments:
>>  args[0] - path to csv input file
>> args[1] - cardinality of the matrix (number of columns)
>> args[2] - path the resulting Mahout's SVD input file
>>
>> Method:
>> The code below, goes over the csv file, and for each matrix column, creates
>> a SequentialAccessSparseVector which contains all the non-zero row entries
>> for this column.
>> Then it appends the column vector to file.
>>
>> Compilation:
>> Copy the java code below into an java file named Convert2SVD.java
>> Add to your IDE project path both Mahout and Hadoop jars. Alternatively, a
>> command line option for compilation is given below.
>>
>>
>> view plain<
>> http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
>> print<
>> http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
>> ?<
>> http://bickson.blogspot.com/2011/02/mahout-svd-matrix-factorization.html#>
>>
>>   1. import java.io.BufferedReader;
>>   2. import java.io.FileReader;
>>   3. import java.util.StringTokenizer;
>>   4.
>>   5. import org.apache.mahout.math.SequentialAccessSparseVector;
>>   6. import org.apache.mahout.math.Vector;
>>   7. import org.apache.mahout.math.VectorWritable;
>>   8. import org.apache.hadoop.conf.Configuration;
>>   9. import org.apache.hadoop.fs.FileSystem;
>>   10. import org.apache.hadoop.fs.Path;
>>   11. import org.apache.hadoop.io.IntWritable;
>>   12. import org.apache.hadoop.io.SequenceFile;
>>   13. import org.apache.hadoop.io.SequenceFile.CompressionType;
>>   14.
>>   15. /**
>>   16.  * Code for converting CSV format to Mahout's SVD format
>>   17.  * @author Danny Bickson, CMU
>>   18.
>>    * Note: I ASSUME THE CSV FILE IS SORTED BY THE COLUMN (NAMELY THE
>> SECOND FIELD).
>>
>>   19.  *
>>   20.  */
>>   21.
>>   22. public class Convert2SVD {
>>   23.
>>   24.
>>   25.         public static int Cardinality;
>>   26.
>>   27.         /**
>>   28.          *
>>   29.          * @param args[0] - input csv file
>>   30.          * @param args[1] - cardinality (length of vector)
>>   31.          * @param args[2] - output file for svd
>>   32.          */
>>   33.         public static void main(String[] args){
>>   34.
>>   35. try {
>>   36.         Cardinality = Integer.parseInt(args[1]);

2011/8/29 myn <[EMAIL PROTECTED]>:
> the best way is to read the sorce code ;

Talking of "view source", has anyone taken a look at these (public
domain) Javascript demos:

http://users.telenet.be/paul.larmuseau/SVD.htm /
http://www.stasegem.be/shop2/SVD.htm

or http://metamerist.com/excanvas/example23a.htm

(I am thinking more for educational usage - interactive tutorials -
than applications)

Could they maybe be useful repackaged, e.g. integrated with
http://sylvester.jcoglan.com/ or http://glmatrix.googlecode.com/ ?

Dan

Funny somebody mentioned the MIT Strang lectures from 99.  I just spent the
weekend watching the first 13.  In case anybody doesn't like being
restricted to watching them in their browser and doesn't have easy access to
download them with iTunesU, you can download them directly from blip (where
academic earth hosts their copies).

Also, if anybody else has trouble maintaining interest through the review, I
recommend trying to different speeds.  I've just been using "mplayer -af
scaletempo -speed 1.5 lecture.m4v".  I took a linear algebra class a few
years ago, but failed to get much out of it because the first few weeks were
a bit too easy and somewhere around week 6 or so I looked up from my french
homework and realized I had no idea what he was talking about.  One of these
days the Universities will catch on and video lectures will revolutionize
the educational landscape.

http://blip.tv/file/get/Aev264-1806Lec1702.m4v
http://blip.tv/file/get/Aev264-1806Lec2300.m4v
http://blip.tv/file/get/Aev264-1806Lec3845.m4v
http://blip.tv/file/get/Aev264-1806Lec4382.m4v
http://blip.tv/file/get/Aev264-1806Lec5103.m4v
http://blip.tv/file/get/Aev264-1806Lec6314.m4v
http://blip.tv/file/get/Aev264-1806Lec7969.m4v
http://blip.tv/file/get/Aev264-1806Lec8818.m4v
http://blip.tv/file/get/Aev264-1806Lec9210.m4v
http://blip.tv/file/get/Aev264-1806Lec10531.m4v
http://blip.tv/file/get/Aev264-1806Lec11902.m4v
http://blip.tv/file/get/Aev264-1806Lec12269.m4v
http://blip.tv/file/get/Aev264-1806Lec13711.m4v
http://blip.tv/file/get/Aev264-1806Lec14407.m4v
http://blip.tv/file/get/Aev264-1806Lec15761.m4v
http://blip.tv/file/get/Aev264-1806Lec16461.m4v
http://blip.tv/file/get/Aev264-1806Lec17485.m4v
http://blip.tv/file/get/Aev264-1806Lec18383.m4v
http://blip.tv/file/get/Aev264-1806Lec19562.m4v
http://blip.tv/file/get/Aev264-1806Lec20378.m4v
http://blip.tv/file/get/Aev264-1806Lec21301.m4v
http://blip.tv/file/get/Aev264-1806Lec22350.m4v
http://blip.tv/file/get/Aev264-1806Lec23986.m4v
http://blip.tv/file/get/Aev264-1806Lec24910.m4v
http://blip.tv/file/get/Aev264-1806Lec24b142.m4v
http://blip.tv/file/get/Aev264-1806Lec25100.m4v
http://blip.tv/file/get/Aev264-1806Lec26244.m4v
http://blip.tv/file/get/Aev264-1806Lec27954.m4v
http://blip.tv/file/get/Aev264-1806Lec28424.m4v
http://blip.tv/file/get/Aev264-1806Lec29320.m4v
http://blip.tv/file/get/Aev264-1806Lec30579.m4v
http://blip.tv/file/get/Aev264-1806Lec31319.m4v
http://blip.tv/file/get/Aev264-1806Lec32881.m4v
http://blip.tv/file/get/Aev264-1806Lec33862.m4v
http://blip.tv/file/get/Aev264-1806Lec34244.m4v
On Mon, Aug 29, 2011 at 6:13 AM, Dan Brickley <[EMAIL PROTECTED]> wrote:

> 2011/8/29 myn <[EMAIL PROTECTED]>:
> > the best way is to read the sorce code ;
>
> Talking of "view source", has anyone taken a look at these (public
> domain) Javascript demos:
>
> http://users.telenet.be/paul.larmuseau/SVD.htm /
> http://www.stasegem.be/shop2/SVD.htm
>
> or http://metamerist.com/excanvas/example23a.htm
>
> (I am thinking more for educational usage - interactive tutorials -
> than applications)
>
> Could they maybe be useful repackaged, e.g. integrated with
> http://sylvester.jcoglan.com/ or http://glmatrix.googlecode.com/ ?
>
> Dan
>