Basic Aggregations¶

Aggregation functions summarize data by computing statistics like sum, mean, and count. They reduce multiple values to a single result.

Column Aggregations¶

Apply aggregations to DataFrame columns.

1. Single Column Mean¶

```python import pandas as pd

df = pd.DataFrame({ 'Age': [25, 30, 35, 40], 'Salary': [50000, 60000, 70000, 80000] })

print(df['Age'].mean()) # 32.5 ```

2. Single Column Sum¶

python print(df['Salary'].sum()) # 260000

3. Multiple Aggregations¶

python print(df['Age'].min()) # 25 print(df['Age'].max()) # 40 print(df['Age'].std()) # 6.45 print(df['Age'].count()) # 4

DataFrame Aggregations¶

Apply aggregations across the entire DataFrame.

1. All Columns Mean¶

python print(df.mean())

Age 32.5 Salary 65000.0 dtype: float64

2. All Columns Sum¶

python print(df.sum())

3. Summary Statistics¶

python print(df.describe())

Age Salary count 4.000000 4.000000 mean 32.500000 65000.000000 std 6.454972 12909.944487 min 25.000000 50000.000000 25% 28.750000 57500.000000 50% 32.500000 65000.000000 75% 36.250000 72500.000000 max 40.000000 80000.000000

Common Aggregation Methods¶

Methods available on Series and DataFrame.

1. Central Tendency¶

python df['Age'].mean() # Arithmetic mean df['Age'].median() # Middle value df['Age'].mode() # Most frequent value

2. Dispersion¶

python df['Age'].std() # Standard deviation df['Age'].var() # Variance df['Age'].sem() # Standard error of mean

3. Quantiles¶

python df['Age'].quantile(0.25) # First quartile df['Age'].quantile([0.25, 0.75]) # Multiple quantiles

Numeric Only Aggregations¶

Handle mixed data types.

1. numeric_only Parameter¶

```python df = pd.DataFrame({ 'Name': ['Alice', 'Bob'], 'Age': [25, 30], 'Salary': [50000, 60000] })

df.mean(numeric_only=True) # Exclude 'Name' ```

2. Select Numeric Columns¶

python df.select_dtypes(include='number').mean()

3. Specific Columns¶

python df[['Age', 'Salary']].mean()

Axis Parameter¶

Aggregate along rows or columns.

1. axis=0 (Default)¶

```python df = pd.DataFrame({ 'A': [1, 2, 3], 'B': [4, 5, 6] })

df.sum(axis=0) # Sum each column ```

A 6 B 15 dtype: int64

2. axis=1¶

python df.sum(axis=1) # Sum each row

0 5 1 7 2 9 dtype: int64

3. Row Mean¶

python df['RowMean'] = df.mean(axis=1)

Handling Missing Values¶

Aggregation methods handle NaN by default.

1. skipna=True (Default)¶

```python import numpy as np

s = pd.Series([1, 2, np.nan, 4]) s.mean() # 2.333... (ignores NaN) ```

2. skipna=False¶

python s.mean(skipna=False) # NaN (includes NaN)

3. Count Non-NaN¶

python s.count() # 3 (only non-NaN values) len(s) # 4 (all values including NaN)

Exercises¶

Exercise 1. Create a DataFrame with columns 'math', 'science', and 'english' containing scores for five students. Use .mean() to compute the average score for each subject (column-wise) and for each student (row-wise with axis=1).

import pandas as pd

df = pd.DataFrame({
    'math': [85, 90, 78, 92, 88],
    'science': [80, 85, 82, 95, 90],
    'english': [75, 88, 91, 84, 79]
})
print("Average per subject:\n", df.mean())
print("\nAverage per student:\n", df.mean(axis=1))

Exercise 2. Given a DataFrame of daily stock prices for three tickers, compute the describe() statistics. Then extract only the 'mean' and 'std' rows from the result using .loc.

import pandas as pd
import numpy as np

np.random.seed(42)
df = pd.DataFrame({
    'AAPL': np.random.uniform(140, 160, 20),
    'MSFT': np.random.uniform(340, 360, 20),
    'GOOGL': np.random.uniform(130, 150, 20)
})
stats = df.describe()
print(stats.loc[['mean', 'std']])

Exercise 3. Create a Series with some NaN values. Demonstrate the difference between .count() and len() by printing both. Then compute the .sum() and show that NaN values are skipped by default but can be included by setting min_count.

import pandas as pd
import numpy as np

s = pd.Series([10, 20, np.nan, 40, np.nan])
print(f"count(): {s.count()}")  # 3
print(f"len(): {len(s)}")       # 5
print(f"sum(): {s.sum()}")      # 70 (NaN skipped)