Exercise: Data Manipulation with Pandas

What you'll learn

• How to import and use pandas for data manipulation
• Techniques for loading and exploring CSV datasets
• Methods for sorting data by single and multiple columns
• How to group data and calculate statistics
• Strategies for filtering data with conditions

:computer: Part One: Quick Quiz

Time estimate: 10 minutes

Instructions:

1. Read each question carefully
2. Choose the correct answer for each question
3. Submit your form when you are done

:information_source: Info :link: Access the Exercise Form Here

:computer: Part 2: Hands-on Coding Activity

Time estimate: 30-40 minutes

:warning: Important Before starting, save a copy of the template to your Google Drive:

:link: Copy the Colab Template Here

Click "File" -> "Save a copy in Drive" to create your own version

Step One: Import Libraries (5 minutes)

1. Open your copied Colab notebook
2. Import pandas and numpy for data manipulation
3. Run the cell to verify imports work correctly

:bulb: If you get an import error, try restarting the runtime from the menu: Runtime -> Restart runtime

Step 2: Loading Your Dataset (5 minutes)

1. Click the folder icon on the left sidebar in Colab
2. Upload your CSV file using the upload button
3. Use pd.read_csv() to read the file into a DataFrame
4. Assign it to a variable (e.g., df) tip Common Challenge If your file has special characters, try adding encoding='utf-8' to your read_csv() function:

df = pd.read_csv('filename.csv', encoding='utf-8')

Step 3: Data Exploration (10 minutes)

1. View the first 5 rows using df.head()
2. Check all column names with df.columns
3. Review data types and info using df.info()
4. Check for missing values with df.isnull().sum()
5. Remove unnecessary columns using df.drop()

:bulb: Helpful Hint To see more rows, use df.head(10) or df.tail() to see the last rows

Step 4: Sorting Data (5 minutes)

1. Sort by a single column in ascending order:
   python

   df.sort_values('column_name')
   

2. Sort by a single column in descending order:
   python

   df.sort_values('column_name', ascending=False)
   

3. Sort by multiple columns with different orders:
   python

   df.sort_values(['col1', 'col2'], ascending=[True, False])
   

Step 5: Grouping Data (10 minutes)

1. Group by a single column and count values:
   python

   df.groupby('column_name').size()
   

2. Calculate sum of numeric values by group:
   python

   df.groupby('category')['numeric_column'].sum()
   

3. Find mean values by group:
   python

   df.groupby('category')['numeric_column'].mean()
   

4. Group by multiple columns:
   python

   df.groupby(['col1', 'col2']).size()
   

5. Sort your grouped results:
   python

   grouped_data.sort_values(ascending=False)
   

:bulb: Tip Use .reset_index() after grouping to convert the result back to a regular DataFrame

Step 6: Filtering Data (5 minutes)

1. Filter by a single condition:
   python

   df[df['column'] > value]
   

2. Filter by multiple conditions using & (and) or | (or):
   python

   df[(df['col1'] > value1) & (df['col2'] == 'value2')]
   

3. Store filtered results in new variables:
   python

   filtered_df = df[df['column'] == 'specific_value']
   

:warning: Remember Always use parentheses around each condition when combining multiple filters!

Troubleshooting Common Issues

DataFrame not showing?

• Make sure to print or display your DataFrame: just type df in a cell

Column not found error?

• Check exact column names with df.columns - they are case-sensitive
• Remove extra spaces in column names: df.columns = df.columns.str.strip()

Memory error with large datasets?

• Try loading only specific columns: pd.read_csv('file.csv', usecols=['col1', 'col2'])

:emoji:️ Submission

:warning: Important - Submit Your Work! Before submitting:

1. Review all your code outputs
2. Make sure all steps are completed
3. Save your Colab notebook (File -> Save)
4. Download a copy for your records (File -> Download -> Download .ipynb)

:link: Submit your completed exercise here

Your submission helps track your progress and understanding of pandas data manipulation!