Methods of calculation of determinants of a higher order
are based on the properties of determinants. Here we consider two methods which being combined together result in the most efficient computing technique.
Before formulating the theorem, let us introduce a few definitions.
Let A be a square matrix of the order n.
By removing the i-th row and the j-th
column, we obtain the submatrix of A, having
the order n - 1.
The determinant of that submatrix is called the minor of the element
ai,j, which is denoted
by Mi,j:
The cofactor of the element ai,j
is defined as the minor Mi,j
with the sign . It is denoted by the symbol Ai,j:
.
The following theorem gives a systematic procedure of determinant calculation.
The determinant of a matrix A equals
the sum of the products of elements of any row of A
and the corresponding cofactors:

|
The above theorem is known as the expansion of the determinant according
to its i-th row.
Since a matrix and its transpose have equal determinants, the theorem
can be formulated in terms of expanding of a determinant by a column:
The determinant of a matrix A equals
the sum of the products of elements of any column of A
and the corresponding cofactors:
|
Due to the theorem, a given determinant of the order n
is reduced to n determinant of the order
(n - 1).
By means of elementary row and column operations on matrices, a determinant
can be transformed to a triangular form, which is easily evaluated.
Let us define the elementary operations. Note that any techniques which
are developed for rows may be also applied to columns.
- Interchange two rows or columns.
As a result, the determinant changes its sign.
- Multiply a row (column) by a nonzero number.
As a consequence of this operation, the determinant is multiplied by
that number.
- Add a row (column) multiplied by a number
to another one.
By this operation, the determinant holds its value.

|