Get autoit to do accurate math

[nullschritt]

It seems that after mutliplying a number so many time over, it becomes impossible to reverse it to it's original value.

It seems this is because autoit automatically does scientific notation, and automatically rounds off the decimal place. (like this 7.222782900955e+50)

Is there any way to make autoit return the WHOLE accurate number?

[AZJIO]

[nullschritt]

Thanks, I am doing this with matrix mathematics though, I have my doubts as to the efficiency of calling this udf to preform the math for every equation(as matrix multiplication, and inversion require a lot of calculations), but I will report the results.

Edit: Yep it has increased the execution time by 10,000+X (had to kill the process it was taking so long)

Changed

For $i = 0 To $n-1
For $j = 0 To $p-1
For $k = 0 To $m-1
$s = $s + $A [$i][$k] * $B [$k][$j]
Next
$C [$i][$j] = $s
$s = 0
Next
Next

to

For $i = 0 To $n-1
For $j = 0 To $p-1
For $k = 0 To $m-1
$s = _BigNum_Mul(_BigNum_Add($s, $A[$i][$k]), $B[$k][$j])
Next
$C [$i][$j] = $s
$s = 0
Next
Next

Any other suggestions, or perhaps a way to do the same thing with matricies (like maybe a bignum_mMult(matrix1, matrix2) )?

Edited August 19, 2012 by nullschritt