time-series-decomposer

Time series decomposition separates data into trend, seasonal, and residual components using additive or multiplicative models. This Skill leverages pandas and statsmodels to break down patterns for analysis, forecasting, and anomaly detection.

Anomaly detection in time series residuals identifies unusual spikes or dips by isolating the leftover noise after removing trend and seasonality. This Skill extracts those residual points to flag operational metrics deviations.

Sales forecasting using decomposition predicts future values by identifying underlying trends and recurring seasonal patterns in historical data. This Skill applies statsmodels to generate basic forecasts from separated components.

Yes, decomposition supports both additive and multiplicative models to handle time series where seasonal variations are constant or proportional to trend levels. You specify the model and seasonal period for accurate component separation.

Time series decomposition requires pandas, numpy, scipy, statsmodels, and matplotlib. These libraries provide the data structures, statistical modeling, and visualization needed to separate and analyze trend, seasonal, and residual components.