This dummy code did not work, from another code that got the result, I wrote this code and it did not work with the true result

@@adrien3995 can you share the code

hola E:\PROGRAMACION\Metodos numericos\Gauss_Jordan.py:20: RuntimeWarning: invalid value encountered in double_scalars a[k,j] /= pivot E:\PROGRAMACION\Metodos numericos\Gauss_Jordan.py:21: RuntimeWarning: invalid value encountered in double_scalars b[k] /= pivot llegue a tener estos problemas me ayudas a solucionarlos porfa ? o sabes como lo puedo hacer?

How

If a system has a non-unique solution, its coefficient matrix cannot be reduced to the identity matrix because the main diagonal will have one or more zero elements. Thus, Gauss-Jordan is used to solve systems with unique solutions. Even though Gauss-Jordan can be used to solve systems with multiple constant terms (b-vectors), they are still unique solutions. I hope I could answer your question. Thanks for the comment.

@@mechtutorcom Thanks for answering. Sir. I would love to know if these line of code can be used to eliminate ANY matrix into a reduced row echelon form. Your matrix is belongs to a unique full rank type. Most of the matrices are not.

@@Troglodyte2021 This code is limited to eliminate full rank matrices for rows and columns, so the matrix should be square since the elimination loops go the same number (n). So, some modifications should be done to handle the number of columns and the locations of non-zero elements to satisfy the conditions of the reduced row echelon form. Thank you for the comment.

@@mechtutorcom hello sir may i know what modification to do?

During the partial pivoting, the code checks if the current pivot element a[k,k] is zero. If 0.0 is used, the extremely small values, like 1.234e-14, can be skipped although they approach zero and can cause large errors in next operations. For that reason, the programmer can select a value which all absolute values of a[k,k] less than that value are considered zero. In this code, I selected 1.0e-12. It depends mainly on the precision of computer variables, the tolerated error and the nature of the problem. In most textbooks of Numerical Analysis, you can find a chapter about this topic because it affects almost all numerical methods. Thank you very much for your comment.

@@mechtutorcom thanks it really helped a lot