type() returns the type of a variable.

isinstance(var, type) check whether var is of type.

Python variables cannot be declared without assigning a value to it.

Boolean

True & False

Boolean values can be treated to numbers. True is 1, False is 0.

Equality: ==.

To test for identity, use is keyword. E.g. var1 is var2 == False. Identity means the two names refer to the same object.

any(iter) returns True if any element of the iterable is true. all(iter) returns True if all elements of the iterable is true.

Numbers

  • int: can be arbitrarily large.
  • float: is accurate to 15 decimal places. int() method truncates towards 0. round(value, ndigits) round towards nearest even digit (1.5 and 2.5 both round to 2), and ndigits can be negative (round to 10, 100, etc.).
  • fractions.Fraction(numerator, denominator). By import fractions, we can use the built-in fraction class. Note denominator cannot be 0.
  • math package. It contains constants math.pi, all trigonometric functions.
  • 0 and 0.0 are False, everything else is True. Be careful about rounding error in floating points!!!
  • complex(real, imag) or 2 + 5j generates complex number. All the usual number operations are supported, and use cmath module (instead of math) if necessary.
  • float('inf'), float('-inf'), and float('nan') creates infinity, negative infinity and NaN. To check whether a value is inf or nan, use math.isinf(a) and math.isnan(a). Note == and is cannot be used to check for nan!

Number Operations

  • /: floating point division, returns float no matter what.
  • //: integer division. It always truncates down (towards negative infinity), so 11//2 == 5, -11//2 == -6.
  • ** power. 11 ** 2 == 121.
  • % remainder.

Accurate Implementation

Floating point can’t accurately represent all base-10 decimals. If needed (especially in finance world), use decimal module.

Number Format

Float

Use built-in format(value, definition) function. definition is a string like [<>^]?width[,]?(.digits)[fe] where width is the width of the output, 0 means no override. < is left-justified, > is right-justitied, and ^ is centered. digits is the number of digits in accuracy. f means float, e or E means exponential notation. Place a , before . inserts thousands separator.

Integer

bin(x), oct(x), and hex(x) converts integer x into its binary, octal, and hexadecimal representation.

format(x, 'b|o|x') does the same thing, except it won’t produce 0b|o|x prefix.

Integer is signed, so negative value has a negative sign prefixing it. To convert to the unsigned representation, add the maximum number to it to set the bit length.

To convert back to decimal, use int(str, base).

None

None is Python’s null.

Comparing None to anything other than None is always False.

None itself is evaluated to False.

Bytes

An immutable sequence of numbers between 0 and 255. To mutate a byte in bytes object, convert it to bytearray object. Indexing on it returns integers, not individual characters.

To create with literals, use b'someString'. To use constructors, use bytes().

String and bytes are different datatypes, so they cannot be concatenated with each other.

Bytes can be converted to strings using by.decode(encoding) method. Similarly, s.encode(encoding) converts a UTF-8 string into bytes.

Bytes supports most of the same built-in operations as text string, but not string formatting. Some of them also works with bytearray too. To use regular expression, pattern needs to be declared as byte string too.