transform Method¶

The transform() method applies a function to each group and returns a result with the same shape as the original DataFrame.

Basic Concept¶

Transform preserves the original index.

1. Difference from agg¶

```python import pandas as pd

df = pd.DataFrame({ 'group': ['A', 'A', 'B', 'B'], 'value': [1, 2, 3, 4] })

agg: returns one value per group¶

df.groupby('group')['value'].mean()

group¶

A 1.5¶

B 3.5¶

transform: returns same-length Series¶

df.groupby('group')['value'].transform('mean')

0 1.5¶

1 1.5¶

2 3.5¶

3 3.5¶

```

2. Same Shape Output¶

Transform output aligns with original DataFrame.

3. Broadcast Group Values¶

Each row gets its group's aggregated value.

LeetCode Example: First Activity Date¶

Find each player's first login date.

1. Sample Data¶

python activity = pd.DataFrame({ 'player_id': [1, 1, 2, 2], 'event_date': pd.to_datetime([ '2024-01-01', '2024-01-02', '2024-01-03', '2024-01-04' ]) })

2. Transform with min¶

python activity["first"] = activity.groupby("player_id")["event_date"].transform('min') print(activity)

player_id event_date first 0 1 2024-01-01 2024-01-01 1 1 2024-01-02 2024-01-01 2 2 2024-01-03 2024-01-03 3 2 2024-01-04 2024-01-03

3. Each Row Gets Group Min¶

Every row for player 1 shows their first date.

Common Use Cases¶

Typical transform applications.

1. Group Normalization¶

python df['normalized'] = df.groupby('group')['value'].transform( lambda x: (x - x.mean()) / x.std() )

2. Percent of Group Total¶

python df['pct_of_group'] = df.groupby('group')['value'].transform( lambda x: x / x.sum() )

3. Rank Within Group¶

python df['group_rank'] = df.groupby('group')['value'].transform('rank')

transform vs apply¶

Key differences between methods.

1. Output Shape¶

```python

transform: must return same shape¶

apply: can return any shape¶

```

2. Function Requirements¶

```python

transform: function must return same-length Series¶

apply: more flexible¶

```

3. Performance¶

```python

transform: often faster for built-in functions¶

apply: more flexible but potentially slower¶

```

Multiple Columns¶

Transform multiple columns at once.

1. Same Function¶

python df[['value1', 'value2']] = df.groupby('group')[['value1', 'value2']].transform('mean')

2. Different Functions¶

```python

Use apply for different functions per column¶

```

3. Preserving Original¶

```python

Create new columns instead of overwriting¶

df['value_mean'] = df.groupby('group')['value'].transform('mean') ```

Exercises¶

Exercise 1. Given a DataFrame with 'department' and 'salary' columns, use groupby().transform('mean') to add a new column 'dept_avg_salary' that shows each department's average salary on every row.

import pandas as pd

df = pd.DataFrame({
    'department': ['IT', 'IT', 'HR', 'HR', 'IT'],
    'salary': [70000, 65000, 50000, 55000, 72000]
})
df['dept_avg_salary'] = df.groupby('department')['salary'].transform('mean')
print(df)

Exercise 2. Use transform to normalize values within each group: for each group, compute (x - mean) / std. Add the result as a 'normalized' column.

import pandas as pd

df = pd.DataFrame({
    'group': ['A', 'A', 'A', 'B', 'B', 'B'],
    'value': [10, 20, 30, 100, 200, 300]
})
df['normalized'] = df.groupby('group')['value'].transform(
    lambda x: (x - x.mean()) / x.std()
)
print(df)

Exercise 3. Use transform('sum') to compute each row's value as a percentage of its group total. Add a column 'pct_of_group' that shows what fraction each row contributes to its group's total.

import pandas as pd

df = pd.DataFrame({
    'region': ['East', 'East', 'West', 'West'],
    'sales': [100, 300, 200, 200]
})
df['pct_of_group'] = df['sales'] / df.groupby('region')['sales'].transform('sum')
print(df)