Solve Problem 3.8.14 with the following modification: instead of using the flops command (which is obsolete in MATLAB and OCTAVE), use the timing functions tic and toc. To time a specific operation, the syntax is

tic; toc;

Youll then be comparing the time taken to do the various operations, not the number of flops.:

Consider the 50 x 50 matrix A' with the main-diagonal entries A 2 for all i, the superdiagonal entries (A') ii + 1-1 for all i, the subdiagonal entries (A')-1,i-1 for all i, d all other entires equal to zero. As in the paragraph following (3.64), view A' as the sum A+ CR in the form where B is the 25 x 25 matrix that looks like a small version of A', where F and Eare 25 x 1 with F--10... o and E [o o O-1] immediately before and after the computation-to find the number of floating-point operations required to do so. (b) Use MATLAB to find the inverse of A' by means of Theorem 3.63(a)-computing the inverse of A by forming the obvious partitioned matrix using the inverse B-1-and compare the numbers of floating-point operations required by the two methods