A friend wanted to know why I enjoy programming in Python so much more than programming in other languages. So I will describe some special features of Python that make it much easier to quickly implement algorithms.
I also made drafts of how these tasks would be solved in most programming languages. When I say "most," I mean most languages that are widely spread (so C/C++/Java is much more important than almost any other languages combined). I know that these tasks would be solved completely differently in functional programming languages.
Rapid, Readable Programming
Intuitive Looping Through Lists
You can loop through every list-like datastructure like this:
for element in list:
print(element)
Arbitrary Integer Size
Task: Print the sum of the digits of $2^{100000}$.
Java:
import java.math.BigInteger;
public class test {
public static void main(String[] args) {
BigInteger a = new BigInteger("2");
a = a.pow(100000);
int sum = 0;
for (int i=0; i < a.toString().length(); i++) {
sum += a.toString().charAt(i);
}
System.out.println(sum);
}
}
Python: (was much faster in both computation and programming time!)
big = 2 ** 100000
sumOfDigits = 0
for digit in str(big):
sumOfDigits += int(digit)
print(sumOfDigits)
Python has no need for a special class as it has arbitrary length integers (see source)
Switch Values of Variables
Task: You want to make sure that variable a is smaller than b ($a < b$).
Most languages:
tmp = a
a = min(a, b)
b = max(tmp, b)
Python:
a, b = min(a, b), max(a, b)
Return More Than One Variable
Task: Evaluate $f: \mathbb{R}^2 \rightarrow \mathbb{R}^3, f(x, y) := (x^2, y^2, x+y)$
Most languages:
double[] function(double x, double y) {
double[] returnValues = new double[3];
returnValues[0] = x * x;
returnValues[1] = y * y;
returnValues[2] = x + y;
return returnValues;
}
double[] values = function(4, 5);
System.out.printf("Part 1: %.2f%n", values[0]);
System.out.printf("Part 3: %.2f%n", values[2]);
Python:
def function(x, y):
return (x * x, y * y, x + y)
a, b, c = function(4, 5)
print(f"Part 1: {a:.2f}")
print(f"Part 3: {c:.2f}")
This is called "tuple unpacking". It actually returns only one variable (a tuple), but creating the tuple is so easy that it doesn't feel like creating another variable.
Short Initialization
Task: Get a string representation of a list from the standard library
Java:
import java.util.Arrays;
import java.util.List;
public class test {
public static void main(String[] args) {
List<Integer> myList = Arrays.asList(1, 3, 3, 7);
System.out.println(myList);
}
}
Python:
my_list = [1, 3, 3, 7]
print(my_list)
Both get the same result.
Chaining Comparisons
Task: You would like to check if $x \in [-5, 42]$.
Most languages:
if (-5 <= x && x <= 42)
Python:
if -5 <= x <= 42:
pass
Enumeration
Task: You have a list and you would like to print it, prefixed with the index in the list.
Java:
List<Integer> myList = Arrays.asList(1, 3, 3, 7);
for (int i = 0; i < myList.size(); i++) {
System.out.printf("%d: %d%n", i, myList.get(i));
}
Python:
my_list = [1, 3, 3, 7]
for nr, element in enumerate(my_list):
print(f"{nr}: {element}")
Named String formatting
Python allows you to give parameters names:
print("The %(foo)s is %(bar)i." % {"foo": "answer", "bar": 42})
any() and all()
Task: You have a very long list and you want to know if a prime is in this list.
Java:
List myList = (List initialisation and assignment of many values)
boolean isPrimePresent = false;
for (int element : myList) {
if (isPrime(element)) {
isPrimePresent = true;
break;
}
}
if (!isPrimePresent) {
System.out.println("The list did not containe a prime.");
}
Python:
my_list = [4, 4, 9, 12]
if not any(is_prime(x) for x in my_list):
print("The list did not contain a prime")
See also: StackOverflow answer from steveha.
Testing and Documentation
Doctest
You can write documentation and unit tests at the same time! Take a look at doctest — Test interactive Python examples.
Sphinx
Documentation can be generated from partially docstrings, partially rst files with Sphinx. It can be uploaded to pythonhosted.org just like neurolab did it (see also sphinx).
The Rest
Lists and Generators
I already wrote an article about Python Lists and Python Generators. I love Pythons lists ☺
for ... else
Task: You have a very long list and you want to know if a prime is in this list.
Java:
List myList = (List initialisation and assignment of many values)
boolean isPrimePresent = false;
for (int element : myList) {
if (isPrime(element)) {
isPrimePresent = true;
break;
}
}
if (!isPrimePresent) {
System.out.println("The list did not contain a prime.");
}
Python:
my_list = [1, 3, 3, 7]
for element in my_list:
if is_prime(element):
break
else:
print("The list did not contain a prime.")
Step Through Lists
Task: Print only every n-th element of an iterable.
for element in my_list[::n]:
print(element)
Dynamically add properties to objects and classes
class Node(object):
value = 3
a = Node()
b = Node()
print(a.value)
""" colorize the node! """
# print(a.color) ==> AttributeError
# thats ok, although the object originally had no attribute "color"
a.color = "white"
print(a.color)
# You can even add a property to the class
Node.special = "here is it"
print(b.special)
Imaginary numbers
Python directly supports usage of imaginary numbers:
(2j + 1) ** 2
Output: (-3+4j)