1. Try to use the cause-effect graph technique to give a Boolean-function definition of a pro-gram that computes the square root of its input value. Why can't you do this? 2. By following the same philosophy we used to reduce the size of the test set for cause-effect graphs, design a criterion that reduces the size of test sets derived by the multiple-condition-coverage criterion given in Exercise 6.6. Exercise 6 Let (Cs) be defined as in Exercise 6.4. The -condition-coverage criterion can be defined as follows: Each test set must make all conditions Cs true and false in all possible ways, based on the values of their constituents. For Stance, if; is cis and co then we must generate four cases that would make cs, true and cs 2 true, cs , false and c s2 true, etc. Give (possibly minimal) test sets satisfying the multiple-condition-coverage criterion for the following program fragment: Clearly, the multiple-condition-coverage criterion is finer than the edge-coverage criterion. Which is finer, the multiple-condition-coverage criterion or the truth-assignment. criterion? Which is finer, the multiple-condition-coverage criterion or the condition-coverage criterion?