So I'm currently working on C code translated from Fortran that finds the inverse of a 17x17 matrix using Guassian Elimination. However, the output has two rows that are tranposed for some strange reason. I'm stuck trying to determine if the indeces being used to increment through the 17x17 matrix to create the indentity matrix are incrementing correctly despite the C to Fortran translation.

Index numbers for matrices in Fortran start at one whereas in C they start at zero. In addition, the way in which C references rows and columns is opposite of Fortran. So where C calls Row then Column with the rightmost index incrementing the fastest, Fortran calls Column then Row with the leftmost index incrementing the fastest. It's a bit of challenge trying to sort through this code in my head.

This entry was posted on 1.10.2005 at 10:54. You can skip to the end and leave a response.