Python’s array module provides typed arrays (all elements must be of the same type).
-> More memory-efficient than lists for large datasets of uniform data types.
-> Unlike lists, arrays cannot mix types (e.g., integers and strings).
from array import array
# Syntax: array(typecode, [elements])
arr = array('i', [1, 2, 3, 4, 5]) # 'i' for integersType Code: Data Type 'i': Signed integer 'f': Float 'd': Double 'u': Unicode char ... ...
->All elements must be of the same type. ->Only 1D Arrays; Cannot create multi-dimensional arrays (e.g., matrices). ->Cannot perform operations on entire arrays without loops. ->Requires type codes (e.g., 'i', 'd'), which can be confusing. ->Less intuitive than NumPy for numerical work.
NumPy (Numerical Python) is the backbone of scientific computing in Python. Here’s why it’s indispensable:
Speed: NumPy is written in C and optimized for performance. Operations on NumPy arrays are much faster than Python lists.
Memory
Efficiency: NumPy arrays store data more compactly than Python lists, saving memory.
Convenience: It provides a vast collection of mathematical functions to operate on arrays without writing loops.
Foundation for Other Libraries: Libraries like Pandas, Matplotlib, and SciPy are built on top of NumPy.
Numpy is not another programming language but a Python extension module.
NumPy is a library for working with arrays and matrices. It allows you to perform mathematical operations on entire arrays without writing loops.
Numerical computations (math, statistics, physics, etc.). Working with large datasets (e.g., images, signals, matrices). Any task involving arrays or matrices.
General-purpose tasks (e.g., storing mixed data types). Small datasets where NumPy’s overhead isn’t justified.
What they are: Built-in Python containers for storing collections of items (e.g., numbers, strings, mixed types). Memory: Stores each item as a separate object → more memory (slower for big data). Speed: Uses Python loops → slower for math operations. Use when: You need flexibility (e.g., mixed data types, small datasets).
import numpy as np
str = ["HI", 13, "Imane"]
print(str)
=> output : ["HI", 13, "Imane"]Technical Explanation
- Python lists store each item as a separate object in memory.
- Each object has extra info (like its type, reference count, etc.).
- When you loop over a list, Python has to fetch each object one by one from different memory locations.
- Elements can be of different types.
What they are: Specialized containers for numerical data (from the numpy library). Memory: Stores data in a single block → less memory (faster for big data). Speed: Uses vectorized operations (no loops) → much faster for math. Use when: You work with numbers, matrices, or large datasets.
a = np.array([10, 5, 60])
print(a)
=> output : [10 5 60]- NumPy arrays store all elements in a single, contiguous block of memory.
- All elements are of the same type (e.g., all int32 or float64).
- No extra overhead for pointers or type info.
- All elements must be of the same type .
Vectorization means applying an operation to entire arrays at once, without writing loops.
NumPy does this by calling optimized C/Fortran code under the hood.
No Python loops = much faster.
-> Python loops are interpreted, meaning Python checks each step (type, memory, etc.).
-> NumPy bypasses Python and uses pre-compiled C code for operations.
import numpy as np
# Without vectorization (slow)
my_list = [1, 2, 3]
squared_list = [x**2 for x in my_list] # Python loop
# With vectorization (fast)
my_array = np.array([1, 2, 3])
squared_array = my_array ** 2 # NumPy does it all at once!each file represent an exercice with my own solution
26:24 - 11.) 29:35 - 30:48 - 33:17 - 34:57 - 40:19 - 16.) How to add a border (filled with 0’s around an existing array? (np.pad) 43:41 - 17.) Evaluate some np.nan expressions 48:32 - 18.) Create a 5x5 matrix with values 1,2,3,4 just below the diagonal 56:01 - 19.) Create an 8x8 matrix and fill it with a checkerboard pattern 1:02:35 - 20.) Get the 100th element from a (6,7,8) shape array 1:07:09 - 21.) Create a checkerboard pattern 8x8 matrix using np.tile function 1:16:22 - 22.) Normalize a random 5x5 matrix 1:24:20 - 23.) Create a custom dtype that describes a color as four unsigned bytes (RGBA) 1:29:27 - 24.) Multiply a 5x3 matrix by a 3x2 matrix (real matrix product) 1:32:54 - 25.) Given a 1D array, negate all elements which are between 3 and 8, in place 1:37:16 - 26.) Default “range” function vs numpy “range” function 1:40:25 - 27.) Evaluate whether expressions are legal or not 1:55:41 - 28.) Evaluate divide by zero expressions / np.nan type casting 1:57:48 - 29.) How to round away from zero a float array? 1:59:22 - 30.) How to find common values between two arrays? 2:00:19 - 31.) How to ignore all numpy warnings? 2:03:24 - 32.) Is np.sqrt(-1) == np.emath.sqrt(-1) ?? 2:05:22 - 33.) Get the dates of yesterday, today, and tomorrow with numpy 2:19:39 - 34.) How to get all the dates corresponding to the month of July 2016? 2:27:27 - 35.) How to compute ((A+B)*(-A/2)) in place (without copy)? 2:35:00 - 36.) Extract the integer part of a random array of positive numbers using 4 different methods 2:40:47 - 37.) Create a 5x5 matrix with row values ranging from 0 to 4 2:43:07 - 38.) Use generator function that generates 10 integers and use it to build an array 2:43:58 - 39.) Create a vector of size 10 with values ranging from 0 to 1, both excluded. 2:48:49 - 40.) Create a random vector of size 10 and sort it. 2:51:07 - 41.) How to sum a small array faster than np.sum? 2:54:37 - 42.) Check if two random arrays A & B are equal 2:58:48 - 43.) Make an array immutable (read-only) 3:02:14 - Puppies are great 3:03:06 - 44.) Convert cartesian coordinates to polar coordinates 3:20:37 - 45.) Create a random vector of size 10 and replace the maximum value by 0 3:23:58 - 46.) Create a structured array with x and y coordinates covering the [0,1]x[0,1] area 3:26:25 - 47.) Given two arrays, X and Y, construct the Cauchy matrix C (Cij = 1/(xi-yj)) 3:34:31 - 48.) Print the min/max values for each numpy scalar type 3:36:50 - 49.) How to print all the values of an array? 3:39:23 - 50.) How to find the closest value (to a given scalar) in a vector?