Transposing a fixed size matrix

Explicit matrix transposition is not often required in Gandalf, because of the extensive support for implicit transpose in other matrix operations. However where necessary, computing the transpose can be achieved using

      Gan_Matrix33 m33A, m33B;

      /* set up m33A using e.g. gan_mat33_fill_q() */
      gan_mat33_tpose_q ( &m33A, &m33B ); /* B = A^T */
      m33B = gan_mat33_tpose_s ( &m33A ); /* B = A^T */
      gan_mat33_tpose_i ( &m33A ); /* A = A^T, result in-place in A */