14

Next week I will make a presentation about mutation test in my class. The professor give me the paper An Experimental Determination of Sufficient Mutant Operators, written by A. Jefferson Offutt et.al. From the paper, I can get a conclusion that the number of mutants generated for a program is roughly proportional to the product of the number of data references times the number of data objects. Here is a simple Fortran-77 program:

FUNCTION Min(I, J)
Min = I
IF (J .LT. I) Min = J
RETURN

The author, i.e. A. Jefferson Offutt, say there are total 44 mutants for the above program. I just don't how to get the 44 mutants. Does anyone know how to calculate the number of data references and the number of the data objects and get the 44 mutants?

Any ideas will be appreciated.

Now, I believe the 44 mutants is not gained just by multiple the number of data references and the number of data objects. Other mutant operators also contribute to the total number of mutants of the above Fortran-77 program. I edit this question in order to prevent other people to keeping answering it. I don't know how to close a question or mark it answered

machinarium
  • 631
  • 6
  • 17
  • That is just a - return the smaller one of the two - function. I haven't read (nor intend to) the given paper, but I doubt your problem is in that part of the code. – Rook Apr 20 '11 at 03:58
  • The paper you have references another paper when making the statement about the 44 mutants. Seems to me you need to read that paper to understand this. – eriktous Apr 20 '11 at 10:37
  • @eriktous, According to my reading, there not reference another paper when making the statement about the 44 mutants. – machinarium Apr 26 '11 at 14:14
  • Now, the 44 mutants are generated by 22 mutant operators including replace the operands and the relational operators. So it is impossible to get 44 just by multiple the number of data references and the number of data objects. – machinarium Apr 27 '11 at 14:59
  • 4
    If you want to close this, then just create an answer yourself, wait 24 hours or so, and then accept it as the right answer. – MJB May 11 '11 at 20:31
  • If you want to close this question, just say so at its end. I can do the first vote to close it. – Roland Illig Jun 11 '11 at 06:58
  • Voting to close as too localized per @machinarium's edit: "I edit this question in order to prevent other people to keeping answering it. I don't know how to close a question or mark it answered". – Jeremy W. Sherman Jun 20 '11 at 16:02
  • Closed this as too localized per OP's edit. – Tim Post Jun 26 '11 at 11:06

0 Answers0