Module 6: Advanced Indexing

Advanced indexing allows you to select non-contiguous elements and filter data based on logical conditions. This is the bedrock of data cleaning and filtering in Python.

---

Lesson 13: Boolean Indexing

Boolean indexing uses a "mask" (an array of True/False values) to pick elements.

How it works:

1. You apply a condition to an array (e.g., arr > 5).
2. This creates a boolean mask of the same shape.
3. You pass that mask back into the array to get only the True values.

Multiple Conditions:

Use & (and), | (or), and ~ (not) for bitwise logic. Note: Parentheses are mandatory when combining conditions.

---

Lesson 14: Fancy Indexing

Fancy indexing refers to passing an array of indices to access multiple elements at once.

Selection with Lists

You can pass a list of specific indices you want to extract. arr[[0, 2, 5]] extracts elements at 0, 2, and 5.

MD Selection

In 2D arrays, you can pass lists for both rows and columns. matrix[[0, 1], [2, 0]] picks (0,2) and (1,0).

---

Mini Project: Filter Large CSV Data

Imagine you have a dataset of 1000 items with random prices and categories.

Challenge: Create an array of 100 random prices (np.random.randint(10, 500, 100)). Filter the prices that are between 100 and 200 using boolean indexing.

Key Takeaways

✅ Boolean indexing is the fastest way to filter data.
✅ Always use parentheses when using multiple boolean conditions.
✅ Fancy indexing allows picking specific, non-contiguous elements.