NumPy: Basics

2. Core Concepts

2.1 🔢 ndarray
1. The main object in NumPy is called an ndarray. ndarray means N-dimensional array. A NumPy array: 👉 Stores many values together 👉 Stores values of the same data type 👉 Has shape, size, and data type

3. Creating NumPy Arrays

3.1 Creating Arrays from Lists and Tuples
1. We can create NumPy arrays from Python lists or tuples. If values have different data types, NumPy may convert them into a common data type. For example, if we mix integers and floats, NumPy converts all values to float. From List import numpy as npmy_list = [1, 2, 3]my_array = np.array(my_list) # We can also use np.asarray(my_list)print(my_array) # [1 2 3] print(type(my_array)) # <class numpy.ndarray> From Tuple import numpy as npmy_tuple = (4, 5, 6)my_array = np.array(my_tuple, dtype=float) print(my_array) # [4. 5. 6.] Creating 2D Array import numpy as npmy_2d_list = [[1, 2], [3, 4]]my_2d_array = np.array(my_2d_list) print(my_2d_array)Output [[1 2] [3 4]]
3.2 Inspecting Array Properties
1. We can inspect NumPy arrays using shape, size, dtype. Example import numpy as nparr = np.array([[1, 2, 5], [3, 4, 7]]) # [[1 2 5] # [3 4 7]] print(arr.shape) # (2, 3) print(arr.dtype) # int64 or int32 print(arr.size) # 6 📌 These properties help us understand the structure of an array.

4. Accessing Elements in NumPy Arrays

4.1 Indexing
1. Indexing means selecting a specific element from an array. Like Python lists, NumPy indexing starts from 0. We can also use negative indexing. Index: 1D Array import numpy as nparr = np.array([10, 20, 30, 40, 50])print(arr[0]) # 10 print(arr[2]) # 30 print(arr[-1]) # 50 print(arr[-2]) # 40 We can also update values: arr[-1] = 80 print(arr) # [10 20 30 40 80] Index: 2D Array import numpy as nparr_2d = np.array([[1, 2, 3],[4, 5, 6],[7, 8, 9] ])print(arr_2d[0]) # First row: [1 2 3] print(arr_2d[1, 2]) # 2nd row, 3rd column: 6 print(arr_2d[-1, -2]) # Last row, 2nd last column: 8 Updating a value: arr_2d[0, 0] = 10 print(arr_2d)Output: [[10 2 3] [ 4 5 6] [ 7 8 9]]

4.2 Slicing
1. Slicing means selecting a part of an array. Syntax: array[start:stop] start is included stop is not included Slicing: 1D Array import numpy as nparr = np.array([10, 20, 30, 40, 50])print(arr[1:4]) # [20 30 40] print(arr[:3]) # [10 20 30] print(arr[2:]) # [30 40 50] print(arr[-3:-1]) # [30 40] For slicing a 2D array, we first slice by row and then by column. The slicing before , is for rows, and the slicing after , is for columns. Example import numpy as nparr_2d = np.array([[1, 2, 3],[4, 5, 6],[7, 8, 9] ])print(arr_2d[0:2]) # Sub-matrix with row 0 and 1 Output: [[1 2 3] [4 5 6]] Select All rows and first two columns: print(arr_2d[:, :2])Output: [[1 2] [4 5] [7 8]] Select rows from index 1 to end and last column: print(arr_2d[1:, -1])Output: [6 9]

5. Adding Elements to NumPy Arrays

5.1 np.insert
1. np.insert() adds values at a specific index. Insert on 1D Array import numpy as nparr = np.array([3, 5, 8, 2, 7])new_arr = np.insert(arr, 2, 10) print(new_arr)Output: [ 3 5 10 8 2 7] For 2D array, we can either add new row or column. axis is used to determine whether to add on row or on column. axis=0 indicates new row has to be added and axis=1 indicates new column has to be added. Example: Add Row in 2D Array import numpy as nparr_2d = np.array([[1, 2], [3, 4]])new_arr = np.insert(arr_2d, 1, [5, 6], axis=0) print(new_arr)Output: [[1 2] [5 6] [3 4]] Example: Add Column in 2D Array import numpy as nparr_2d = np.array([[1, 2], [3, 4]])new_arr = np.insert(arr_2d, 1, [5, 6], axis=1) print(new_arr)Output: [[1 5 2] [3 6 4]]

5.2 np.append
1. np.append() adds values at the end of an array. For 2D array similar to np.insert, we use axis parameter to control whether to add at last row or add at last column. For 2D arrays, the new row or column should also be given in 2D format. Append: 1D Array import numpy as nparr = np.array([1, 2, 3]) new_arr = np.append(arr, [4, 5]) print(new_arr)Output: [1 2 3 4 5] Append: Add Row in 2D Array import numpy as nparr_2d = np.array([[1, 2], [3, 4]]) new_arr = np.append(arr_2d, [[5, 6]], axis=0)print(new_arr)Output: [[1 2] [3 4] [5 6]] Example: Add Column in 2D Array import numpy as nparr_2d = np.array([[1, 2], [3, 4]]) new_arr = np.append(arr_2d, [[5], [6]], axis=1) print(new_arr)Output: [[1 2 5] [3 4 6]]

6. Removing Elements

6.1 np.delete
1. np.delete() removes values from an array. We can specify the index which needs to be dropped. For 2-D, axis=0 drops row and axis=1 drops column. Example: 1D Array import numpy as nparr = np.array([10, 20, 30, 40, 50]) new_arr = np.delete(arr, 2) print(new_arr)Output: [10 20 40 50] Example: Remove Row from 2D Array import numpy as nparr_2d = np.array([[1, 2], [3, 4], [5, 6]]) new_arr = np.delete(arr_2d, 1, axis=0) print(new_arr)Output: [[1 2] [5 6]] Example: Remove Column from 2D Array import numpy as nparr_2d = np.array([[1, 2], [3, 4], [5, 6]]) new_arr = np.delete(arr_2d, 0, axis=1) print(new_arr)Output: [[2] [4] [6]]

7. Vectorized Operations

7.1 Array with Constants
1. Vectorized operations apply an operation to every element of an array. This means we do not need to write loops. Common operations include: +, -, *, /, **, //, %, <, <=, >, >=, ==, !=Example import numpy as nparr = np.array([1, 2, 3, 4])print(arr + 10) # [11 12 13 14] print(arr * 2) # [2 4 6 8] print(arr > 2) # [False False True True] print((arr % 2 == 0) & (arr > 2)) # [False False False True] 📌 NumPy applies the operation to all elements at once.
7.2 Array with Array
1. When two arrays have the same shape, NumPy applies operations element-wise. Example import numpy as nparr1 = np.array([1, 2, 3]) arr2 = np.array([4, 5, 6])print(arr1 + arr2) # [5 7 9] print(arr1 * arr2) # [ 4 10 18] print(arr1 > arr2) # [False False False]
7.3 Matrix Multiplication
1. For matrix multiplication, use @. import numpy as npmat_a = np.array([[1, 2], [3, 4]]) mat_b = np.array([[5, 6], [7, 8]])print(mat_a @ mat_b)Output: [[19 22] [43 50]] 📌 Important: * performs element-wise multiplication, but @ performs matrix multiplication.

8. Mathematical Functions

8.1 Applying Math Functions
1. NumPy provides many mathematical functions. These functions usually work element-wise. Example import numpy as npa1 = np.array([2, 5.2, 8.7])print(np.sqrt(a1)) # Square root print(np.power(a1, 2)) # Square print(np.sin(a1)) # Sine value print(np.round(a1, 2)) # Round to 2 decimals print(np.abs(a1)) # Absolute value print(np.radians(a1)) # Degrees to radians print(np.degrees(a1)) # Radians to degrees 📌 NumPy math functions can work on the whole array at once.

9. String Functions: np.char

9.1 Applying String Functions
1. NumPy also supports string operations using np.char. These functions work element-wise on string arrays. Example import numpy as nps1 = np.array(["hi ", " all"])print(np.char.upper(s1)) # ["HI " " ALL"] print(np.char.title(s1)) # ["Hi " " All"] print(np.char.strip(s1)) # ["hi" "all"] print(np.char.replace(s1, " ", "_")) # ["hi_" "_all"] print(np.char.split(s1)) # [["hi"] ["all"]] print(np.char.find(s1, "h")) # [0 -1] print(np.char.count(s1, "l")) # [0 2] print(np.char.str_len(s1)) # [3 4]

10. Aggregate Functions

10.1 What are Aggregate Functions?
1. Aggregate functions summarize data. They help answer questions like: ❓What is the total? ❓What is the average? ❓What is the minimum? ❓What is the maximum? We can aggregate the whole array, or aggregate by rows and columns using the axis parameter. Example: Whole Array import numpy as npa2 = np.array([[1, 2], [3, 4]])print(np.sum(a2)) print(np.mean(a2)) print(np.median(a2)) print(np.std(a2)) print(np.var(a2)) print(np.min(a2)) print(np.max(a2)) print(np.argmin(a2)) print(np.argmax(a2)) print(np.clip(a2, 2, 3)) print(np.cumsum(a2)) print(np.ptp(a2)) print(np.quantile(a2, 0.5))Example: Along Axis import numpy as npa2 = np.array([[1, 2], [3, 4]])print(np.sum(a2, axis=0)) # Column-wise sum [4 6] print(np.mean(a2, axis=1)) # Row-wise mean [1.5 3.5] 📌 Aggregate functions help summarize data quickly.
10.2 Axis Meaning
1. For a 2D array: axis=0 means column-wise operation ↓. axis=1 means row-wise operation →. 📌 Axis controls the direction of calculation.

11. Applying Custom Functions to NumPy Arrays

11.1 np.vectorize
1. np.vectorize() is a convenient way to apply our own function to every element of a NumPy array. Example import numpy as npdef can_have_license(age):return age >= 16vectorized_license_check = np.vectorize(can_have_license) ages = np.array([10, 16, 18, 12]) result = vectorized_license_check(ages) print(result) # [False True True False] 📌 np.vectorize() helps apply a custom function to each value in an array.

12. NumPy Constants

12.1 Common NumPy Constants
1. NumPy provides some useful mathematical constants.

12.2 Example
1. import numpy as npprint(np.pi) print(np.inf) print(-np.inf) print(np.nan) 📌 These constants are useful in math, statistics, and data cleaning.