It's important to keep the knowledge of CS basics alive. That way you're not caught off guard when new opportunities arise. This notebook will contain questions that I encountered from Leetcode, HackerRank and other places as well as concepts that I've forgotten that need studying.

Hopefully this will also be good material for blog posts and revisions when needed.

Each directory will contain a subject. If there is no directory for a certain subject then I haven't studied this subject at all.

Python is neat. This repo is 100% Python, so naturally I picked up some really great tools. Some I've used in the past and got re-introduced to as I'm building this repo.

zip is used to combine iterables into tuples.

a = [1, 3]
b = [2, 4]
c, d = zip(a, b)

assert(c == (1, 2))
assert(d == (3, 4))

The reason why this is great to use for algorithm questions is because it allows you to quickly split arrays of the same size into usable iterables.

A clean way to do this is using the * syntax as follows.

intervals = [[10, 20], [1, 5], [50, 90]]
start_times, end_times = zip(*intervals)

assert(start_times == (10, 1, 50))
assert(end_times == (20, 5, 90))

Very clean :D

all is a clean way to verify if an iterable is all True. Basically all is a reducer that just uses the and operator.

all([True, True, True])  # True
all([True, True, False])  # False

This could be a good way for finding out if something is sorted or not.

any is a clean way to verify if an iterable has any True values. Basically any is a reducer that just uses the or operator.

all([True, True, True])  # True
all([False, False, True])  # True
all([False, False, False])  # False

array = [1, 5, 6, 7, -1]
n = len(array)
all([array[i-1] <= array[i] for i in range(1, n - 1)])

There is no replacing sorted as a utility. sorted uses a method called Timsort which uses a hybrid of mergesort and insertion sort with space O(N).

You can modify sorted by giving it a key function with a lambda. For example, if you want to sort elements that are of type TreeNode you can extract the val in the lambda itself.

enumerate is a simple function that gives a more convenient way of looping an array. Instead of writing a typical for i in range... it gives you the index and the element.

array = [1, -3, 5, -4]
for index, element in enumerate(array):
    print(index, element)

# Will return
# 0 1
# 1 -3
# 2 5
# 3 -4

math.inf is a convenient way to get a number that is bigger than any other number. This is a great way to start off min or max when you don't know what the range is for the input.

import math
high = -math.inf
low = math.inf

array = [1, 2, 4, 5, 6]
for elem in array:
    high = max(high, elem)
    low = min(low, elem)

Some graph questions have a tedious iteration especially in an adjacency matrix question. Instead of having a bunch of if conditions where each one moves a pointer once, a cleaner and less error-prone way would be to put all the places that the pointers will go and then iterate through them.

def visit_perpendicular(graph, i, j):
    """ Visit perpendicular nodes to node [i][j]"""
    for ii, jj in [(i + 1, j), (i - 1, j), (i, j + 1), (i, j - 1)]:
        # do something
        graph[ii][jj] = 2

defaultdict allows you to define a special type of library that whenever accessed with a key that doesn't exist, it initializes a value for it. This way you don't get any KeyError exceptions as you're coding.

defaultdict takes a callable not a value for it to be called during initialization. For example, you can use defaultdict(list) to return a new list everytime the dict is accessed. This especially useful for building graphs.

from collections import defaultdict

def build_graph(edges):
    graph = defaultdict(list)
    for node, neighbour in edges:
        graph[node].append(neighbour)
    return graph

Generators in python are a great way to keep going without storing into a list.

Simple way to reverse a list or a string. Doesn't modify the original list.

l = [1, 2, 3, 4, 5]
l = l[::-1]  # [5, 4, 3, 2, 1]

You can get the byte size of anything in Python by simply encoding the string in utf-8 and get the length.

def byte_size(string):
    return len(string.encode('utf-8'))

byte_size('?')  # 1
byte_size('Hello World')  # 11