If denotes the transpose of the matrix , where and are integers such that , then the number of such matrices for which is
Condition means that is an orthogonal matrix, i.e. . In an orthogonal matrix, every row (and every column) is a unit vector that is orthogonal to the other two rows (or columns).
Because all entries are integers, a unit vector can contain only ; moreover exactly one entry must be and the others must be . (Otherwise its squared length would exceed .) Thus every row of is a “signed” standard basis vector.
Write the given matrix as with . Since the non-zero entries can only be we already have .
Step 1: Orthogonality of row 1 and row 2
Row 1 is , Row 2 is . Their dot product must vanish: , so .
Step 2: Orthogonality of row 1 and row 3
Row 3 is . Dotting with Row 1 gives , hence .
Step 3: Orthogonality of row 2 and row 3
Row 2 is now . Their dot product is , so .
Step 4: Unit-length of each row
Row 1 already has squared length (since ).
Row 2 has squared length (since ).
Row 3 has squared length (since ).
Thus all rows are unit vectors and pairwise orthogonal, satisfying orthogonality completely.
After these deductions the matrix must be
Step 5: Counting the matrices
Each of can independently take two values ( or ). Hence the total number of admissible matrices is .
Therefore the required number of matrices is .
Option C which is:
Tagged Counting matrices with given properties. Machine-assigned, so read it as a strong shortlist rather than a verdict.