Lære Expanding Window Features

Sveip for å vise menyen

Expanding window statistics are a powerful tool for extracting cumulative patterns from time series data. Unlike rolling windows, which calculate metrics over a fixed number of recent observations, expanding windows compute statistics from the start of the series up to the current point. This approach is especially useful when you want to capture long-term trends or overall progress, such as cumulative sums, means, or other aggregations that reflect all available information up to each time step. Expanding window features are commonly used in financial analysis to monitor running totals or averages, in process monitoring to track ongoing performance, and in any scenario where the cumulative effect is more meaningful than short-term fluctuations.


              1234567891011121314151617
            
import pandas as pd

# Create a sample time series
data = {'value': [10, 12, 13, 15, 18, 20]}
index = pd.date_range('2023-01-01', periods=6, freq='D')
series = pd.Series(data['value'], index=index)

# Calculate expanding mean and expanding sum
expanding_mean = series.expanding().mean()
expanding_sum = series.expanding().sum()

print("Original Series:")
print(series)
print("\nExpanding Mean:")
print(expanding_mean)
print("\nExpanding Sum:")
print(expanding_sum)

Alt var klart?

Takk for tilbakemeldingene dine!

Seksjon 1. Kapittel 5

Spør AI

Spør om hva du vil, eller prøv ett av de foreslåtte spørsmålene for å starte chatten vår

Seksjon 1. Kapittel 5