Data Manipulation with Pandas

While NumPy is excellent for numbers, Pandas is the ultimate library for handling tabular data (rows and columns), similar to working with SQL tables or Excel sheets.

---

19.2.1 Series vs. DataFrames

Pandas has two primary data structures:

• Series: A single column of data.
• DataFrame: A two-dimensional table composed of multiple columns (Series).

       DataFrame
   ┌─────────┬──────────┐
   │ Product │ Price    │ ───> Column (Series)
   ├─────────┼──────────┤
   │ Laptop  │ 50000    │
   │ Mouse   │ 1000     │
   └─────────┴──────────┘

---

19.2.2 Basic Operations

Installing Pandas:

pip install pandas

Creating and Mutating DataFrames:

import pandas as pd

# 1. Create a DataFrame manually
data = {
    "Product": ["Laptop", "Mouse", "Keyboard"],
    "Price": [50000, 1000, 2000],
    "Quantity": [2, 10, 5]
}
df = pd.DataFrame(data)
print("Initial DataFrame:")
print(df)

# 2. Add a calculated column
df["Total_Cost"] = df["Price"] * df["Quantity"]
print("\nWith Total Cost:")
print(df)

Output:

Initial DataFrame:
    Product  Price  Quantity
0    Laptop  50000         2
1     Mouse   1000        10
2  Keyboard   2000         5

With Total Cost:
    Product  Price  Quantity  Total_Cost
0    Laptop  50000         2      100000
1     Mouse   1000        10       10000
2  Keyboard   2000         5       10000

---

19.2.3 Filtering and Aggregations

One of the main strengths of Pandas is the ease of filtering rows and aggregating column values:

# 1. Filtering: Get products with a price greater than 1500
expensive_items = df[df["Price"] > 1500]
print("Expensive items:")
print(expensive_items)

# 2. Aggregation: Calculate summaries
print("Total revenue:", df["Total_Cost"].sum()) # Output: 120000
print("Average quantity:", df["Quantity"].mean()) # Output: 5.66