Exercise 2: Bubble Sorting

• You are asked to read in, and sort, a list of numbers which are said to be of type double. double is yet another C data type - like int, float and char. double is most similar to type float: it is also used to approximately represent "real" numbers. It differs from type float is that it normally has greater precision (more significant digits) and greater range (can handle larger and smaller numbers - can have a bigger absolute value in the "exponent" part in "scientific" notation). Type double is actually the most common "preferred" type for general purpose calculations. By that I mean that, if you need to represent numbers, your first choice should normally be to use type double; and you should use something else only where there is some definite reason for doing so.

  Having said that, type double does introduce a few new problems of its own. Therefore I suggest that you initially ignore the reference to type double. That is, first try to get the program to read and sort a list of numbers of type float (or even of type int). When you have got this working successfully then go back and try to modify it for type double.

• The length of the list of numbers is not fixed in advance. But just as with MATMUL, you need to fix the size of array for holding the list to some definite value, which will then be the maximum length the program can deal with. The exercise does not explicitly state a required maximum. I suggest that you get it to work for lists at least up to 100 numbers long.

• The original BUBBLE.C program does not read in the array to be sorted. Instead the list is explicitly initialised in the program. However, in the modified program, the list must be read in at run time. You have already encountered code for doing that general sort of job in program MATMUL.C, so look at that for examples or inspiration.