Programming Forums - Bizarre TI-89 matrix algebra error (not language specific)

I am trying to write a Basic program for my TI-89 calculator to solve matrices. For some weird reason this function is not built-in to the calculator, which has stuff to do mostly everything else.

I don't think this is language-specific; I think that if you know other programming languages you may be able to help. I haven't done BASIC in five years, and this is my first go at TI-89 BASIC, and it didn't take me very long to start from scratch. If you know about matrixes and you can program then you probably can help.

The matrix I am trying to do is
1 -1 0 5
-1 1 5 2
0 1 1 0

The solution of this matrix is
1 0 0 18/5
0 1 0 -7/5
0 0 1 7/5

But my function, solvemat(x), spits this out instead
1 1/4 0 13/4
0 5/4 0 -7/4
0 -1/4 1 7/4

So far you're thinking, he just made some stupid error in his code that he'll catch in a minute before I even reply. But I don't think so.
Point 1: Applying the function twice, solvemat(solvemat(x)) returns the correct result
Point 2: The key value, where it seems to be screwing up, is on the number at position [2,2] (indexing in BASIC starts from 1). Altering that number in any way causes the program to function perfectly correctly. If I put .5 or 2 instead of 1 in that position, or even 1.0001 or 9999/10000 instead of 1, it gives the right answer.

WHAT THE HELL IS GOING ON?

Here is my code, with Java-style comments (the TI-89 comment character is unprintable):

solvemat(x)           // the name of the function.  x is the matrix

Func                   //stating that it is a function and not a program

Local j, L, i, k, o  // These are integers.  Note:  

                //Sorry about the non-

                //descriptive names, I had to type on an 

                //alphabetic non-qwerty 

                     //keyboard so I didn't want them to be long

// j and L will be dealt with shortly

// i is the loop variable for columns

// k is the loop variable for rows

// o keeps track of the last row that properly starts with a 1



colDim(x) - 1 -> j  //colDim(x) = number of columns in x.  -> j 

                   // means "assign to j"

rowDim(x) -> L           //rowDim(x) = number of rows in x

0->o             //o is initialized to 0

For i, 1, j, 1  // for(i = 1; i <= j; i++)

        o+1 -> k

        While x[k, i] = 0

                k + 1 -> k

                If k > L

                        Goto down //I know it's lame but there

        EndWhile                //is no "continue" in TI-89 basic



        mRow(1/(x[k, i]), x, k) -> x



                /* this built-in matrix solving function means 

                   "multiply row k of matrix x

                   by 1/(x[k, i]) and put the result back into

                   matrix x." What I am doing here is putting

                   a 1 at the beginning of the row */



        rowSwap(x, k, o+1) // swaps row k and row o+1



        o+1 -> o        //I could have done this a line earlier

                        // and had neater code, but I didn't

 

/* Now I have a row that is in the proper location, and it has

 a 1 at the beginning of it.  Now I will use that row to

 eliminate all the nonzero elements of the matrix above and below 



the 1. */



        For k, 1, L, 1 // for(k = 1; k <= L; k++) also L = 

                        // rowDim(x) from earlier



                If k = o // that's an 'o'

                        Goto d2



                If x[k, i] != 0

                        mRowAdd(-x[k, i], x, o, k) -> x



        /* This built-in matrix solving function means

        "multiply row o of matrix x by -x[k, i] and add the

        result to row k, and the resulting matrix goes back

        into matrix x */



                Lbl d2

        EndFor

        Lbl down

EndFor



Return x



EndFunc

Please help! It's driving me nuts. I suppose I could just make another function solvemat2(x) that is just solvemat(solvemat(x)) but I'm doing this for fun anyway and I want to know what's happening.

Maybe it's a calculator bug.