NumPy: Transforming Data

2. np.where() for Conditional Logic

2.1 🤔 What is np.where()?
1. np.where() helps us apply if-else logic on NumPy arrays. It can be used in two ways: 1. Find indices where a condition is true 2. Return different values based on a condition
2.2 Syntax
1. np.where(condition, value_if_true, value_if_false) Interpretation: 👉 If condition is True, use first value. 👉 Otherwise, use second value.
2.3 💻 Examples
1. import numpy as nparr = np.array([1, 2, 3, 4, 5])Example 1: Find index positions result = np.where(arr > 3)print(result)Output: (array([3, 4]),) 📌 This means values greater than 3 are found at index 3 and index 4. Example 2: Create categories category = np.where(arr > 3, "High", "Low")print(category)Output: ['Low' 'Low' 'Low' 'High' 'High'] Example 3: Modify values conditionally height = np.array([150, 160, 170, 180, 190])new_height = np.where(height <= 160, height + 2, height)print(new_height)Output: [152 162 170 180 190] 📌 Simple idea: np.where() is like applying if-else to every value in an array.

3. np.select() for Multiple Conditions

3.1 🧭 What is np.select()?
1. np.select() is used when we have multiple conditions. It works like an if → elif → elif → else chain.
3.2 Syntax
1. np.select(condition_list, choice_list, default=value)condition_list — List of conditions choice_list — Output for each condition default — Output when no condition matches
3.3 💻 Example
1. import numpy as nparr = np.array([-1, 2, 3, 4, 5])conditions = [(arr > 0) & (arr < 2),(arr >= 2) & (arr < 4),arr >= 4 ]choices = ["Low", "Medium", "High"]result = np.select(conditions, choices, default="Invalid")print(result)Output: ['Invalid' 'Medium' 'Medium' 'High' 'High'] 📌 Simple idea: Use np.where() for simple if-else. Use np.select() for multiple conditions.

4. Reshape, Resize, Flatten, Ravel, and Transpose

4.1 🔄 Why Transform Array Shape?
1. Sometimes we need to change the shape of data. Example: [1 2 3 4 5 6] can be changed into: [[1 2 3] [4 5 6]] Shape transformation helps us prepare data for analysis, visualization, and Machine Learning.

4.3 reshape()
1. reshape() changes the shape of an array without changing the data. The new shape must have the same total number of elements. Example import numpy as nparr = np.array([[1, 2], [3, 4], [5, 6]])reshaped_arr = np.reshape(arr, (2, 3))print(reshaped_arr)Output: [[1 2 3] [4 5 6]] Original array remains unchanged: print(arr)Output: [[1 2] [3 4] [5 6]] Invalid reshape print(np.reshape(arr, (4, 2))) This gives an error because the original array has 6 elements, but shape (4, 2) needs 8 elements. Using -1 NumPy can automatically calculate one missing dimension using -1. print(np.reshape(arr, (-1, 2)))Output: [[1 2] [3 4] [5 6]] Flatten using reshape: print(np.reshape(arr, (-1,)))Output: [1 2 3 4 5 6] 📌 Simple idea: reshape() changes the layout only when the number of elements fits.

4.4 resize()
1. np.resize() returns a new array with the requested shape. If the new size is larger, NumPy repeats values. If the new size is smaller, NumPy removes extra values. Example import numpy as nparr = np.array([[1, 2], [3, 4], [5, 6]]) Resize to larger shape: resized_arr = np.resize(arr, (2, 4))print(resized_arr)Output: [[1 2 3 4] [5 6 1 2]] Resize to same number of elements: resized_arr = np.resize(arr, (2, 3))print(resized_arr)Output: [[1 2 3] [4 5 6]] Resize to smaller shape: resized_arr = np.resize(arr, (2, 2))print(resized_arr)Output: [[1 2] [3 4]] 📌 Simple idea: resize() can repeat or remove values to fit the new shape. 📌 Think back to the egg tray — reshape() rearranges the same eggs into a different tray shape. resize() can also add or remove eggs to fit the new tray.

4.5 flatten()
1. flatten() converts a multi-dimensional array into a 1D array. It returns a copy. Example import numpy as nparr = np.array([[1, 2], [3, 4], [5, 6]])flattened_arr = arr.flatten()print(flattened_arr)Output: [1 2 3 4 5 6] Original array remains unchanged: print(arr)Output: [[1 2] [3 4] [5 6]] 📌 Simple idea: flatten() gives a separate 1D copy.
4.6 ravel()
1. ravel() also converts an array into 1D. It usually returns a view, meaning changes may affect the original array. Example import numpy as nparr = np.array([[1, 2], [3, 4], [5, 6]])raveled_arr = np.ravel(arr)print(raveled_arr)Output: [1 2 3 4 5 6] Modify raveled array: raveled_arr[0] = 99print(arr)Output: [[99 2] [ 3 4] [ 5 6]] 📌 ravel() can be connected to the original array.

4.7 transpose()
1. transpose() swaps rows and columns. Example import numpy as nparr = np.array([[1, 2], [3, 4], [5, 6]])transposed_arr = np.transpose(arr)print(transposed_arr)Output: [[1 3 5] [2 4 6]] Shortcut: print(arr.T) 📌 Simple idea: Transpose changes rows into columns and columns into rows.

5. Merging Arrays

5.1 🧩 Why Merge Arrays?
1. Merging means combining multiple arrays into one. We can combine arrays: 👉 Vertically, by adding rows 👉 Horizontally, by adding columns

5.3 np.concatenate()
1. np.concatenate() combines arrays along a selected axis. Example import numpy as nparr1 = np.array([[1, 2], [3, 4]]) arr2 = np.array([[5, 6], [7, 8]])Combine row-wise result = np.concatenate((arr1, arr2), axis=0)print(result)Output: [[1 2] [3 4] [5 6] [7 8]] Combine column-wise result = np.concatenate((arr1, arr2), axis=1)print(result)Output: [[1 2 5 6] [3 4 7 8]] 📌 Simple idea: axis=0 adds rows. axis=1 adds columns.
5.4 np.vstack()
1. np.vstack() stacks arrays vertically. It adds one array below another. import numpy as nparr1 = np.array([[1, 2], [3, 4]]) arr2 = np.array([[5, 6], [7, 8]])result = np.vstack((arr1, arr2))print(result)Output: [[1 2] [3 4] [5 6] [7 8]] 📌 vstack = stack boxes on top of each other.
5.5 np.hstack()
1. np.hstack() stacks arrays horizontally. It adds one array beside another. import numpy as nparr1 = np.array([[1, 2], [3, 4]]) arr2 = np.array([[5, 6], [7, 8]])result = np.hstack((arr1, arr2))print(result)Output: [[1 2 5 6] [3 4 7 8]] 📌 hstack = place boxes side by side.

6. Copying Arrays

6.1 📋 Assignment vs Copy
1. Assigning one array to another variable does not create a new copy. It creates another name for the same array.
6.2 Reference Example
1. import numpy as nparr = np.array([1, 2, 3])arr_ref = arrarr_ref[0] = 99print(arr)Output: [99 2 3] 📌 Why did this happen? arr_ref and arr point to the same array.
6.3 Copy Example
1. To create an actual copy, use np.copy() or arr.copy(). Example import numpy as nparr = np.array([1, 2, 3])arr_copy = np.copy(arr)arr_copy[0] = 99print(arr) print(arr_copy)Output: [1 2 3] [99 2 3] 📌 Simple idea: Use .copy() when you want to change the new array without changing the original.

7. Linear Algebra Operations

7.1 🧮 What is Linear Algebra in NumPy?
1. NumPy provides functions for matrix operations. These are useful in: 👉 Machine Learning 👉 Data Science 👉 Computer graphics 👉 Statistics 📌 Determinant and inverse come from linear algebra — you do not need to master the math behind them right now to use these functions correctly.

7.3 Matrix Multiplication
1. import numpy as npa = np.array([[1, 2], [3, 4]]) b = np.array([[5, 6], [7, 8]])result = a @ bprint(result)Output: [[19 22] [43 50]]
7.4 Determinant
1. import numpy as npa = np.array([[1, 2], [3, 4]])det = np.linalg.det(a)print(det)Output: -2.0
7.5 Inverse
1. import numpy as npa = np.array([[1, 2], [3, 4]])inverse = np.linalg.inv(a)print(inverse)Output: [[-2. 1. ] [ 1.5 -0.5]]
7.6 Solving Linear Equations
1. Suppose we have: x + y = 52x + y = 8 We can write it as: a = [[1, 1], [2, 1]]b = [5, 8] Then solve: import numpy as npa = np.array([[1, 1], [2, 1]]) b = np.array([5, 8])solution = np.linalg.solve(a, b)print(solution)Output: [3. 2.] This means: x = 3 y = 2 📌 NumPy can solve equations quickly using matrices.