Day 54: Pandas — Excel Controlled by Code!

What Is Pandas?

Open Excel. Click a cell. Type a formula. Click another. Filter manually. Sort manually. One file at a time. Maximum one million rows. That's manual and limited. Now imagine controlling that entire spreadsheet with one line of code. Filter 10 million rows instantly. Sort the whole table in one expression. Merge two datasets automatically. That's Pandas — Excel coming alive through code. No mouse. No clicking. Just power!

Your First DataFrame

import pandas as pd

data = {
    "name"  : ["Rohith", "Sneha", "Arjun"],
    "score" : [87, 92, 45],
    "city"  : ["Hyderabad", "Mumbai", "Delhi"]
}

df = pd.DataFrame(data)
print(df)

Output: A clean table with name, score and city columns. Dictionary keys become column headers. Lists become rows. One call converts your dictionary into a fully structured programmable spreadsheet — your DataFrame!

Exploring Your DataFrame

print(f"Shape   : {df.shape}")
print(f"Columns : {list(df.columns)}")
print(df.head())

df.shape tells you rows and columns — always check this first. df.columns gives you every column header. df.head() peeks at the first 5 rows. These three are your first move every time you open any dataset. Know your data before you touch it!

Real World Connection

When a data scientist at Zomato analyzes 50 million orders, they load it into a DataFrame. When IPL analysts find the best batting average, they use Pandas on match data. When a bank detects fraud patterns, they filter transaction DataFrames. When Netflix decides which shows to produce, they analyze viewing DataFrames with millions of rows. Every data job in every company starts with pd.DataFrame() and ends with insights!

What DataFrames Can Do

# Filter rows — instant on millions of rows
passing = df[df["score"] > 50]

# Sort entire table in one line
sorted_df = df.sort_values("score", ascending=False)

# Get one column
names = df["name"]

print(passing)
print(sorted_df)

df[df["score"] > 50] filters in one expression — no loop needed. sort_values() sorts the entire table by any column instantly. df["name"] extracts one full column. These three operations alone replace hours of manual Excel clicking!

Common Mistakes

Mistake 1 — Wrong import style.

import pandas
df = pandas.DataFrame(data)    # WRONG — every tutorial expects pd!

import pandas as pd
df = pd.DataFrame(data)        # CORRECT — pd is the universal alias!

Mistake 2 — Building DataFrame from a list instead of a dictionary.

df = pd.DataFrame(["Rohith", "Sneha", "Arjun"])  # WRONG — no column headers!

df = pd.DataFrame({"name": ["Rohith", "Sneha", "Arjun"]})  # CORRECT — keys become headers!

Mini Challenge

Quick Quiz

Q: How do you check the number of rows and columns in a DataFrame? A: df.shape — returns (rows, columns) as a tuple!

Q: How do you filter a DataFrame for scores above 50? A: df[df["score"] > 50] — boolean filter, same vector thinking as NumPy!

Q: Why must you always import pandas as pd? A: pd is the universal alias — every codebase, tutorial and company uses it!

Key Takeaways