Question
Let be a non-singular matrix. If column
is replaced by a and that the resulting matrix is called
along with
, then state the necessary and sufficient condition for
to be non-singular.
Properties for the Solution
Since B is a non-singular matrix composed of column vectors, it satisfies the following properties:
1. B is a square matrix with size .
2. Transpose of B will not change its non-singularity nature or rank.
3. B has full rank ⟹ rank(B)=rank()=r=n
4. is a linearly independent system of rank n.
Solution
Now, we can say that BT is a matrix with row vectors and say, column vectors
.
Row vectors can be written as:
As per question, . This means that,
Assume that is a scalar quantity. Our aim is to replace
with a. Let’s say we did that.
Using elementary row operations, if we subtract a with etc., a will become a zero vector. If that happens, the rank of
will become
. This will cause the system to be linearly dependent & lose its non-singular nature.
This can be avoided if is never equal to zero. In that case, the term
will remain non-zero within the system after elementary row operations.
Thus, the necessary & sufficient condition for to be non-singular is
.