[Fwd: Re: archive of the discussion]


Sender: Bruce Miller <bruce.miller@NIST.GOV>
Subject: Re: archive of the discussion


Michal Rehak wrote:
> Greetings,
> I would like to hear some news about JAMA as it is my favorite package
> for matrix computation.
> 
> Recently I have had opportunity to compare API of JAMA against others
> matrix processing packages and libraries under JAVA. I was hoping to
> start using jlapack as it is supposed to be the fastest one. But as I
> saw the "plain" structure with prehistoric organisation, I just decided
> to stay with JAMA. Maybe I got just too used to it. Anyway, JAMA is great.

Nice to hear!

> And the main reason I decided to write this mail for:
> I am sorry to say, that the link at
> [http://math.nist.gov/javanumerics/jama/] to discussion archive
> [http://www.nist.gov/itl/div896/emaildir/jama/maillist.html] is broken.
> I hope the mailing discussion group works.

Yes it does.
The archive seems to have moved to
http://cio.nist.gov/esd/emaildir/lists/jama/maillist.html

I'll update thge jama page to fix that; Thanks for the report.

> And also I would like to bring some suggestions. After a couple of
> months of using JAMA v.1.0.1 a made some modifications in order to
> improve some basic matrix handling. When I found out that the 1.0.2
> version is out, I updated. To spare work on implementing these
> modifications into future updates I rewrote my modifications into a
> static extra class.

The project isn't so much "dead" as "done".
I'll leave it to other project folks to comment on the potential
addition.

> This class contains:
> 
> --------------------------------------------------------------------------
> 
>    /*
>     * ------------------------ Factories ------------------------
>     */
> 
>    /**
>     * Construct a (square) diagonal matrix from a one-dimensional array
>     *
>     * @param vect One-dimensional array of doubles.
>     */
>    public static Matrix createDiag(double[] vect)
> 
> 
>    /**
>     * Construct a column vector (N x 1 Matrix) from a one-dimensional array
>     *
>     * @param vect One-dimensional array of doubles.
>     */
>    public static Matrix colVector(double[] vect)
> 
> 
>    /*
>     * ------------------------ Public Methods ------------------------
>     */
> 
>    /**
>     * Matrix vertical joinig.
>     *
>     * @return join (B under A)
>     * @param A top matrix
>     * @param B bottom matrix
>     * @exception ArrayIndexOutOfBoundsException
>     *                vertical join impossible
>     */
>    public static Matrix joinVertical(Matrix A, Matrix B)
> 
> 
> 
>    /**
>     * Matrix horizontal joinig.
>     *
>     * @return join (B next to A from right)
>     * @param A left matrix
>     * @param B right matrix
>     * @exception ArrayIndexOutOfBoundsException horizontal join impossible
>     */
>    public static Matrix joinHorizontal(Matrix A, Matrix B)
> 
> 
>    /**
>     * Matrix rows removing.
>     * @param A matrix some rows to be eaten from
>     * @param BadRows one dimensional array of integers, containing
> indexes of rows
>     *            to be removed
>     * @return thinned matrix
>     */
>    public static Matrix rowsEater(Matrix A, int[] BadRows)
> 
> 
>    /**
>     * Matrix columns removing.
>     * @param A matrix some columns to be eaten from
>     * @param BadCols one dimensional array of integers, containing
> indexes of cols
>     *            to be removed
>     * @return thinned matrix
>     */
>    public static Matrix colsEater(Matrix A, int[] BadCols)
> 
> 
>    /**
>     * dot product of vectors (A.B)
>     *
>     * @param A must be a column vector
>     * @param B must be a column vector of equal size to A
>     * @return scalar (double) value
>     */
>    public static double dot(Matrix A, Matrix B)
> 
> 
>    /**
>     * cross product of vectors (A x B)
>     *
>     * @param B must be a vector of size (3 rows, 1 column)
>     * @param B must be a vector of equal size as A
>     * @return vector with crossproduct
>     */
>    public static Matrix cross(Matrix A, Matrix B)
> 
> 
>    /**
>     * @param matrix (if not square, the diagonal will be brought from
> the square part only)
>     * @return diagonal of Matrix (n x 1)
>     */
>    public static Matrix diag(Matrix A)
> 
> 
>    /**
>     * counts a square-rooted matrix
>     * @param A origin matrix
>     */
>    public static Matrix sqrt(Matrix A)
> 
> --------------------------------------------------------------------------
> 
> 
> If there is any of interest, I will provide full source. I believe that
> keeping this project alive.. if it isn't dead by now :-( ..is a very
> possitive thing.
> 
> Thank You for response.
> 
> 
> 
> Best regards,
> Michal Rehak
> 
> 
> 
> 
> 
> 


-- 
bruce.miller@nist.gov
http://math.nist.gov/~BMiller/







Date Index | Thread Index | Problems or questions? Contact list-master@nist.gov