Data Manipulation

This lesson explains how to modify and transform data in Pandas DataFrames. It covers operations such as adding and deleting columns and rows, modifying values, reordering data, and other techniques for manipulating tabular information.

Adding and Deleting Columns

New columns can be added by assigning values directly or deleted using drop().

df = pd.DataFrame({
    'name': ['Anna', 'Bernat', 'Carla'],
    'age': [25, 30, 22]
})

# Add a new column
df['city'] = ['Barcelona', 'Girona', 'Lleida']

# Add a calculated column
df['age_doubled'] = df['age'] * 2

# Delete a column
df = df.drop('age_doubled', axis=1)

# Delete multiple columns
df = df.drop(['city'], axis=1)

Direct assignment creates new columns. The drop() method with axis=1 deletes columns. By default, drop() returns a copy, so you need to reassign the result or use inplace=True.

Adding and Deleting Rows

Rows can be added with concat() or loc, and deleted with drop().

# Add a row with loc
df.loc[3] = ['David', 35]

# Add multiple rows with concat
new_rows = pd.DataFrame({
    'name': ['Eva', 'Francesc'],
    'age': [28, 33]
})
df = pd.concat([df, new_rows], ignore_index=True)

# Delete a row by index
df = df.drop(0)

# Delete multiple rows
df = df.drop([1, 2])

The concat() method joins DataFrames vertically. The argument ignore_index=True resets the index numbering.

Modifying Values

Values can be modified using indexing or specific methods.

# Modify a specific value with loc
df.loc[3, 'age'] = 36

# Modify multiple values
df.loc[df['name'] == 'Eva', 'age'] = 29

# Modify with at (faster for single values)
df.at[3, 'name'] = 'Daniel'

# Replace values with replace
df['name'] = df['name'].replace('Daniel', 'David')

The loc and at methods allow assigning new values. The replace() method substitutes specific values throughout the column.

Renaming Columns and Indexes

The rename() method allows changing column and index names.

# Rename columns
df = df.rename(columns={'name': 'nombre', 'age': 'edad'})

# Rename indexes
df = df.rename(index={3: 'three', 4: 'four'})

# Rename multiple elements
df = df.rename(columns={'nombre': 'name'}, index={'three': 3})

The columns parameter accepts a dictionary mapping old names to new names. The same applies to index.

Reordering Columns and Rows

Columns and rows can be reordered by changing the order or sorting the values.

df = pd.DataFrame({
    'age': [25, 30, 22],
    'name': ['Anna', 'Bernat', 'Carla'],
    'city': ['Barcelona', 'Girona', 'Lleida']
})

# Manually reorder columns
df = df[['name', 'age', 'city']]

# Sort rows by a column
df = df.sort_values('age')

# Sort in descending order
df = df.sort_values('age', ascending=False)

# Sort by multiple columns
df = df.sort_values(['city', 'age'])

# Sort by index
df = df.sort_index()

The sort_values() method sorts rows according to the values of one or more columns. The sort_index() method sorts by the row index.

Jordi Petit
Lliçons.jutge.org
© Universitat Politècnica de Catalunya, 2026