Advanced Pandas Techniques: Grouping, Aggregation, Window Functions, and More

This guide presents a series of Pandas examples that illustrate common data‑analysis operations.

1. Grouping aggregation

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

grouped = df.groupby('Category').agg({'Value': ['sum', 'mean']})
print(grouped)

2. Conditional aggregation

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

grouped_filtered = df.groupby('Category').filter(lambda x: x['Value'].mean() > 35)

grouped = grouped_filtered.groupby('Category').agg({'Value': ['sum', 'mean']})
print(grouped)

3. Window function

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

rolling_mean = df['Value'].rolling(window=2).mean()

df['Rolling Mean'] = rolling_mean
print(df)

4. Multi‑index

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

multi_index_df = df.set_index(['Category', df.index])
print(multi_index_df)

5. Melt (reshape)

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

melted_df = pd.melt(df, id_vars=['Category'], value_vars=['Value'])
print(melted_df)

6. Broadcasting

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

df['Value'] += 10
print(df)

7. Concatenation

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

df2 = pd.DataFrame({
    'Category': ['A', 'B'],
    'Value': [70, 80]
})

concatenated = pd.concat([df, df2], ignore_index=True)
print(concatenated)

8. Data merging (join)

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

df2 = pd.DataFrame({
    'Category': ['A', 'B'],
    'Value': [70, 80]
})

joined = df.merge(df2, on='Category', how='left')
print(joined)

9. Time‑series creation

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

df['Date'] = pd.date_range(start='2024-01-01', periods=len(df), freq='D')
print(df)

10. Missing‑value handling

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

df.loc[3, 'Value'] = None

df.fillna(0, inplace=True)
print(df)

11. Custom function application

import pandas as pd

data = {
    'Category': ['A', 'B', 'A', 'B', 'A', 'B'],
    'Value': [10, 20, 30, 40, 50, 60]
}

df = pd.DataFrame(data)

def add_one(x):
    return x + 1

df['Value'] = df['Value'].apply(add_one)
print(df)

Each code block produces the corresponding output shown later in the article, illustrating how Pandas can be used for a wide range of data‑manipulation tasks.