Pandas Fundamentals: 40 Questions to Go From List to DataFrame

Forty questions, one skill ladder — Series -> DataFrame -> inspect -> filter -> sort -> handle nulls -> group — so your muscle memory survives a live interview.

What does the 40-question ladder build?

Part 1 Series/DataFrame -> head/shape/dtypes -> iloc/loc -> filtering & string ops -> sorting -> missing -> groupby -> merging, each mapped to a YouTube reference (Series Basics, Filtering & Sorting, loc vs iloc). Pair with the Cheatsheet for desk reference and the Advanced masterclass after you finish.

Ingredients: 40 prompts from a starter DataFrame (Name, Age, City) expanding to salary and department extensions.

How do you create and inspect a DataFrame?

Setup:

import pandas as pd
import numpy as np

# Q1-Q2: Series vs DataFrame
s = pd.Series([10,20,30,40], index=['a','b','c','d'])
print(s)
df = pd.DataFrame({'Name':['Alice','Bob','Charlie'], 'Age':[25,30,35], 'City':['NY','LA','SF']})
print(df)
 
# Q3-Q6: Inspect
print(df.head(2))
print(df.shape)        # (3, 3)
print(df.columns.tolist())
print(df.dtypes)
print(df.info())
print(df.describe())

Rendered output: s shows labeled index a->d; df three rows with Name object, Age int64. The info/describe block mirrors your EDA opening per EDA playbook.

How do you select, filter, and sort?

# loc vs iloc
print(df.loc[0, 'City'])   # SF via label
print(df.iloc[0, 2])       # SF via position — same cell, different path
 
# Filtering — pay attention to parentheses and &
print(df[df['Age']>26])
print(df[(df['Age']>26) & (df['City']=='NY')])
 
# Strings
print(df[df['Name'].str.contains('li', case=False)])
print(df[df['Name'].str.startswith('A')])
 
# Sorting
print(df.sort_values('Age', ascending=False))
print(df.sort_values(['City','Age'], ascending=[True, False]))

Rendered output: str.contains('li') matches Alice; startswith A isolates her as well — handy for prefix scans like email domains.

How do you handle missing data and group?

# Inject a NaN to demonstrate
df2 = df.copy()
df2.loc[1, 'Age'] = np.nan
print(df2.isna().sum())
print(df2.dropna())
print(df2['Age'].fillna(df2['Age'].median()))
 
# Groupby — team equivalent in master workbook
data = {'Team':['Red','Red','Blue','Blue'], 'Points':[10,15,20,25]}
dfg = pd.DataFrame(data)
print(dfg.groupby('Team')['Points'].mean())
print(dfg.groupby('Team').agg(avg=('Points','mean'), cnt=('Points','count')))

How do you continue?

Re-run every drill on the evolving frame until each answer is one line from memory. Then take the 50Q Master Workbook for grouped aggregation reps and wire the workflow into your Excel pivot dashboard for stakeholder delivery.

Download the Notebook and Practise

This article is a walkthrough of a runnable Jupyter notebook. Download the original .ipynb and run it locally or on Colab — every code block above appears in order.

Continue your track: Data Analyst Roadmap · Python and Pandas Guide · SQL NULL Handbook · SQL JOIN Fan-Out · Topfolio Practice · Data Analyst vs Engineer

Dataset generators where applicable are in courses/workbooks/generators/ — see citations atop for the exact *.py source for this notebook.

Frequently Asked Questions

What is the difference between Series and DataFrame?

Series is one column (1-D with an index); DataFrame is a table of many Series (2-D). df['Age'] returns a Series; df[['Age']] returns a DataFrame.

When do you use loc vs iloc?

loc is label-based (df.loc[0, 'City'] by index label), iloc is position-based (df.iloc[0, 2] by integer position). Mixups cause silent wrong-row bugs.

How do you read a CSV safely?

pd.read_csv('file.csv', dtype={'id': str}, parse_dates=['date'], na_values=['', 'NA']). Specify dtypes to avoid silent int->float coercion.

What does df.describe() actually compute?

Count, mean, std, min, 25%, 50%, 75%, max for numeric columns — a one-line equivalent of five SQL aggregates.