Type Conversion

Data often arrives as the wrong type: numbers read from a file or typed by a user are strings, and calculations may need an int where you have a float. Python converts between types explicitly, by calling the type like a function — int(), float(), str() and so on.

Converting to numbers

PY to_number.py

print(int("42"))
print(int("  -7 "))
print(float("3.75"))
print(float("1e3"))
print(int(9.99))
print(int(True), float(False))

Output

int() and float() accept strings with surrounding whitespace, and int() on a float truncates the decimal part (it doesn't round).

Invalid conversions raise errors

PY invalid.py

print(int("3.5"))

Output

Traceback (most recent call last):
  File "main.py", line 1, in <module>
    print(int("3.5"))
          ~~~^^^^^^^
ValueError: invalid literal for int() with base 10: '3.5'

Unlike JavaScript, which would return NaN, Python stops with a ValueError. To convert "3.5" to an integer, go through float first: int(float("3.5")). When the input comes from a user, catch the error (lesson 41):

PY safe.py

def to_int(text, default=None):
    try:
        return int(text)
    except ValueError:
        return default

print(to_int("12"), to_int("twelve"), to_int("", 0))

Output

12 None 0

Other bases

PY bases.py

print(int("ff", 16))
print(int("1010", 2))
print(int("0x1A", 0))
print(hex(255), bin(10), oct(8))

Output

255
10
26
0xff 0b1010 0o10

The second argument to int is the base. Base 0 means "work it out from the prefix".

Converting to strings

PY to_string.py

print(str(42) + " apples")
print(str(3.0), str(None), str([1, 2]))
print(repr("hi"), repr(3.0))

Output

42 apples
3.0 None [1, 2]
'hi' 3.0

str() gives a readable form for users; repr() gives an unambiguous form for developers (note the quotes around 'hi'). For mixing values into text, f-strings are usually simpler than str() + concatenation (lesson 10).

Converting to bool

PY to_bool.py

print(bool(0), bool(42))
print(bool(""), bool("False"))
print(bool([]), bool([0]))
print(bool(None))

Output

False True
False True
False True
False

Empty values — 0, "", empty collections and None — are false; everything else is true. Note that the string "False" is true because it isn't empty. Lesson 11 covers truthiness fully.

Converting between collections

PY collections.py

letters = list("hello")
print(letters)
print(tuple(letters))
print(sorted(set(letters)))
print("".join(letters))
print(dict([("a", 1), ("b", 2)]))
print(list(range(5)))

Output

['h', 'e', 'l', 'l', 'o']
('h', 'e', 'l', 'l', 'o')
['e', 'h', 'l', 'o']
hello
{'a': 1, 'b': 2}
[0, 1, 2, 3, 4]

Converting to a set removes duplicates, but sets have no order — that's why the example sorts it before printing. Going from a list of characters back to a string uses "".join(...), not str().

Implicit conversion is limited

Python converts automatically only between numeric types, where no information is lost in a surprising way:

PY implicit.py

print(3 + 2.5)
print(True + True)
print(10 * 1.0)

Output

5.5
2
10.0

Mixing an int and a float gives a float. Anything else — like adding a number to a string — raises a TypeError, which prevents a whole class of bugs.

Try it yourself

PreviousNumbers NextStrings