Resampling¶

Resampling changes the frequency of time series data, either downsampling (e.g., daily to monthly) or upsampling (e.g., monthly to daily).

Basic Resampling¶

Change time series frequency.

1. Monthly Average¶

```python import pandas as pd

s = pd.Series( range(100), index=pd.date_range('2025-01-01', periods=100, freq='D') )

monthly = s.resample('M').mean() print(monthly) ```

2. Weekly Sum¶

python weekly = s.resample('W').sum()

3. Quarterly Max¶

python quarterly = s.resample('Q').max()

Common Frequencies¶

Resampling frequency strings.

1. Time-based¶

python s.resample('D').mean() # Daily s.resample('W').mean() # Weekly s.resample('M').mean() # Monthly s.resample('Q').mean() # Quarterly s.resample('Y').mean() # Yearly

2. Business Frequencies¶

python s.resample('B').mean() # Business day s.resample('BM').mean() # Business month end

3. Intraday¶

python s.resample('H').mean() # Hourly s.resample('T').mean() # Minute

OHLC Aggregation¶

Financial price data aggregation.

1. Open-High-Low-Close¶

```python prices = pd.Series( [100, 101, 99, 102, 98, 103], index=pd.date_range('2025-01-01', periods=6, freq='D') )

ohlc = prices.resample('W').ohlc() print(ohlc) ```

2. Standard for Financial Data¶

OHLC is standard for representing price bars.

3. With Volume¶

```python

For DataFrame with price and volume¶

df.resample('W').agg({ 'price': 'ohlc', 'volume': 'sum' }) ```

Aggregation Functions¶

Apply various aggregations when resampling.

1. Built-in Functions¶

python s.resample('M').mean() s.resample('M').sum() s.resample('M').first() s.resample('M').last() s.resample('M').count()

2. Multiple Functions¶

python s.resample('M').agg(['mean', 'std', 'min', 'max'])

3. Custom Function¶

python s.resample('M').apply(lambda x: x.max() - x.min())

Upsampling¶

Increase frequency (requires filling).

1. Daily to Hourly¶

python daily = pd.Series([100, 101, 102], index=pd.date_range('2025-01-01', periods=3, freq='D')) hourly = daily.resample('H').ffill() # Forward fill

2. Fill Methods¶

python s.resample('H').ffill() # Forward fill s.resample('H').bfill() # Backward fill s.resample('H').asfreq() # No fill (NaN)

3. Interpolation¶

python s.resample('H').interpolate()

Practical Examples¶

Financial analysis with resampling.

1. Monthly Returns¶

```python import yfinance as yf

aapl = yf.download('AAPL', start='2023-01-01', end='2024-01-01') monthly_avg = aapl['Close'].resample('M').mean() print(monthly_avg) ```

2. Rolling Analysis on Resampled Data¶

python monthly_returns = aapl['Close'].resample('M').last().pct_change()

3. Plotting¶

```python import matplotlib.pyplot as plt

monthly_avg.plot(title='AAPL Monthly Average Closing Price') plt.show() ```

Exercises¶

Exercise 1. Write code that resamples daily data to monthly frequency using .resample('ME').mean(). Explain what 'ME' means.

Exercise 2. Explain the difference between downsampling and upsampling. Give an example of each.

Exercise 3. Write code that resamples from daily to weekly, computing both the mean and the sum using .resample('W').agg(['mean', 'sum']).

Exercise 4. Create hourly data and upsample to minute-level frequency using .resample('min').ffill().