1. Divide the program into blocks, set breakpoints, and adjust one block at a time, and the next block will succeed.

2. For each block, first declare all declarations and all references of all float type variables in the operation process of this block with int, such as float a->; Int a is just a simple example here, and the actual process is very complicated. If a variable is not corrected somewhere, it is likely to have problems. It must be noted here that the accuracy of the converted data can be insufficient, but the numerical value can not be wrong (more importantly, the array can not exceed the range indicated by int. If it exceeds, it can be changed to other types, or it can be reduced first and then enlarged. )

3. The method called by the object must be changed to int type.

4. According to the size of the shift operation value, pay special attention to the multiplication and division method, and it is easy to exceed the range of int (beyond the range, the value may not be 0, and the value may be a non-zero error value).

5. Calibrate the values and complete the conversion.