Type conversion and debugging

A common surprise: a program asks for a number, the user types 2, and the program prints 22 instead of 4. Why? Because input() returns the string '2', and '2' * 2 repeats the string, giving '22'. Converting first, int('2') * 2, gives 4.

Explicit conversion (type casting) is done by the programmer, using functions:

• int(x) makes a whole number: int('42') is 42. It can lose information: int(9.99) is 9 (the fraction is dropped, not rounded).
• float(x) makes a decimal number: float('2.5') is 2.5.
• str(x) makes a string. You must use it to join a number to a string: 'Total: ' + str(250) works, but 'Total: ' + 250 gives an error.
• chr(x) gives the character for a code: chr(66) is 'B'.
• ord(x) gives the code for a character: ord('B') is 66.

Implicit conversion (coercion) is done by Python itself. In 5 + 2.5, Python changes 5 into 5.0 and gives 7.5, a float. It always converts to the wider type, so that no information is lost.

Debugging means finding and removing errors (bugs). There are three kinds:

• Syntax errors break the rules of the language, like a missing bracket in print(7 + 11. The program does not run at all.
• Logical (semantic) errors let the program run, but give the wrong output. To average 10 and 12, writing 10 + 12 / 2 gives 16.0; the right expression is (10 + 12) / 2, giving 11.0.
• Runtime errors stop a program that looked fine, while it is running: for example, dividing by zero, or int() given text that is not a number.