Documentation Index
Fetch the complete documentation index at: /llms.txt
Use this file to discover all available pages before exploring further.
Automatic and fixed-order (seasonal) ARIMA and autoregression with exogenous regressors, fitted series by series: orders, coefficients with standard errors, AIC/AICc/BIC, forecasts with intervals; unit-root and seasonal-strength tests that choose the differencing.
tsforecast_arima: 2 tools.
| Tool | What it does | Notes |
|---|---|---|
tsforecast_arima | Fits ARIMA-family models to one series (y) or every series of a panel (data): automatic seasonal ARIMA (stepwise AICc search after unit-root tests), ARIMA with given (p,d,q)(P,D,Q) orders, drift, Box-Cox and fixed coefficients, or autoregression on chosen lags; numeric exogenous columns become regressors (give future_exog). | |
tsforecast_differencing | The differencing the automatic ARIMA would choose for each series: d from repeated KPSS stationarity tests (null: stationary) and D from the STL seasonal strength (seasonally differenced when it exceeds 0.64), with the seasonal strength itself (0 none to 1 pure seasonality). |
fincept_describe_tool.