For example, integer data takes up 4 bytes and floating-point data takes up 8 bytes.

If you output with %f, it is read according to 8 bytes of floating-point type, and the natural result will be different.

Sometimes the value is uncertain, sometimes it is 0 or 0.00000.