Defines
#define	MAL_MATRIX(name, type)
	Defines a matrix named name and of type type.
#define	MAL_MATRIX_DIM(name, type, nb_rows, nb_cols)
	Defines a matrix named name of type type, with size (nb_rows, nb_cols ). It can be use for declaration.
#define	MAL_MATRIX_RESIZE(name, type, nb_rows, nb_cols)
	Resize the matrix named name to (nb_rows, nb_cols). This acts as a function.
#define	MAL_MATRIX_NB_ROWS(name)
	Returns the number of rows of matrix name.
#define	MAL_MATRIX_NB_COLS(name)
	Returns the number of cols of matrix name.
#define	MAL_MATRIX_CLEAR(name)
	Clear the matrix name , i.e. put all the value to 0. In some case, it also mean that the data is deallocated. It is wise to resize the matrix after the call to this macro. This macro acts as a function.
#define	MAL_INVERSE(name, inv_matrix, type)
	Returns in inv_matrix, the inverse of matrix name. name should be a square matrix. $inv\_matrix = name^{-1}$ For implementation issue, the type of the data must be specified in type. You should read the documentation of the specific implementation to handle properly the numerical issues. This macro acts as a function.
#define	MAL_PSEUDOINVERSE(matrix, pinv_matrix, type)
	Returns in pinv_matrix the pseudo-inverse of matrix name. name can be of arbitrary size. $inv\_matrix = matrix^{+}$ For implementation issue, the type of the data must be specified in type. This macro acts as a function.
#define	MAL_RET_TRANSPOSE(matrix)
	Returns the transpose of matrix, i.e. go through a temporary matrix.
#define	MAL_TRANSPOSE_A_in_At(matrix)
	Puts the transpose of matrix directly inside the argument.
#define	MAL_RET_A_by_B(A, B)
	Returns the product of A by B, i.e. ${\bf A B}$ This forces the use of a temporary matrix structure.
#define	MAL_C_eq_A_by_B(C, A, B)
	Puts in C the result of the product of A by B, i.e. ${\bf C} = {\bf AB}$ .
#define	MAL_MATRIX_SET_IDENTITY(matrix)
	Set the matrix named matrix to the identity matrix. When matrix is not square, the identity matrix is set to the upper left greatest square matrix included into matrix. The other values are set to zero.
#define	MAL_MATRIX_FILL(matrix, value)
	Fill the matrix named matrix with the value value.
#define	MAL_RET_MATRIX_DATABLOCK(matrix)
	Returns a pointer on the data stored for the matrix.
#define	MAL_MATRIX_C_eq_EXTRACT_A(C, A, type, top, left, nbrows, nbcols)
	This is used to extract a submatrix C from A.
#define	MAL_MATRIX_RET_DETERMINANT(name, type)
	This macro returns the determinant of matrix name.

#define	MAL_VECTOR(name, type)
	declare a vector named name with type type
#define	MAL_VECTOR_DIM(name, type, nb_rows)
	declare a vector named name with type type and of size nb_rows
#define	MAL_VECTOR_SIZE(name)
	returns the size of the vector named name
#define	MAL_VECTOR_RESIZE(name, nb_rows)
	Resizes of the vector named name . It acts as a function.
#define	MAL_VECTOR_FILL(name, value)
	Fill the vector named name with the value value. The user is responsible to make sure that the type are compatible. It acts as a function.
#define	MAL_VECTOR_NORM(name)
	Returns the two norms of the vector named name. More precisely if the vector name is of size , the norm is defined as: $\sqrt{\sum_{i=0}^{i=n-1}x_i^2}$ .
#define	MAL_VECTOR_3D_CROSS_PRODUCT(res, v1, v2)
	Returns in res the cross product of two vectors $v_1,v_2 \in \mathbb{R}^3$ . The cross product $res = v_1 \times v_2$ is defined as: $\begin{eqnarray} res^x &= v_1^y v_2^z - v_2^y v_1^x \\ res^y &= v_1^z v_2^x - v_2^z v_1^x \\ res^z &= v_1^x v_2^y - v_2^x v_1^y \\ \end{eqnarray}$ .
#define	MAL_RET_VECTOR_DATABLOCK(name)
	Returns a direct access to the data information. It is implementation dependent. However most of the time, it is assume to be an array of data.

Defines