'''

Tried to create gaussian elimination as python code

Eventually foudn a script on a site that would do it out.

The website promptly stopped working around 4pm today, but it was from the umbc math department

'''

def gausselim1(themat):
    for i in range(len(themat)):
        for j in range(i+1, len(themat)):
            m = themat[j][i]/themat[i][1]
            themat[j] = [themat[j][k] - m*themat[i][k] for k in range(len(themat[0]))]
    return themat

# wanted ot check it
check_matrix = [
    [2,3,4],
    [6,7,8],
    [11,6,9]]
gausselim1(check_matrix)

'''
I tested it against a 2 by 2 matrix - worked fine.

check1 = [
    [2,4],
    [5,2]]
gausselim1(check1)
>>[[1.0, 2.0], [0.0, -8.0]]

Which is correct.

Then against a 3 by 3 (see the object check_matrix in the code above).  No luck.

The result was -

>>
[  2.0000  3.0000  4.0000]
[  2.0000  1.0000  0.0000]
[ 13.6667  0.0000 -5.6667]

Yes, I'm using a formatting script to change it into something that's neater,
as opposed to the list of lists that it would give.  No, it's not the formatting
script that has the problem - *that* works fine.

I did the gaussian elimination by hand - here are the steps I got
R1 (for Row 1), etc -

R1/2 ->
[1 3/2 2]
[6 7 8]
[11 6 9]

R2 - 6R1 ->
[1 3/2 2]
[0 -2 -4]
[11 6 9]

R3 - 11R1 ->
[1 3/2 2]
[0 -2 -4]
[0 -21/2 -13]

R2/-2 ->
[1 3/2 2]
[0 1 2]
[0 -21/2 -13]

R3 + (21/2)R2 ->
[1 3/2 2]
[0 1 2]
[0 0 8]

R3/8 ->
[1 3/2 2]
[0 1 2]
[0 0 1]

I need someone to -

a) check that the steps I did out by hand are correct -
I've gone over them, but it's late and
it's possible my eyes are playing tricks on me -

if I made an error it would be on the third step (R3-11R1).  I've checked and double checked,
but there was a moment where I ran out of space on the scrap of paper I was working on,
and...well, it would be there if it was anywhere.

b) check the code to tell me where it's going wrong

'''