Time-series API

The public stochx.timeseries namespace is the main applied-analysis surface of StochX.

Data and workfiles

  • TimeSeries — named/indexed univariate series.
  • Workfile — named series, estimation sample and EViews-inspired expression/equation interface.
  • Expression, evaluate — lag/lead/transformation expressions.

Correlation and correlograms

  • acf, pacf
  • correlogram
  • format_correlogram_table, format_correlogram
  • interpret_correlogram

Stationarity

  • adf, dickey_fuller, dickey_fuller_sequential
  • kpss_test, phillips_perron
  • difference, trend_regression, classify_ts_ds

Decomposition and smoothing

  • moving_average, weighted_moving_average
  • exponential_smoothing, holt, holt_winters
  • decompose, seasonal_difference, seasonal_indices
  • fisher_seasonality_test

Model estimation

  • fit_ar, fit_ma, fit_arma, fit_arima, fit_sarima
  • estimate
  • ols, RegressionResult
  • parse_error_terms, ErrorProcess

Box–Jenkins

  • identify_box_jenkins
  • estimate_box_jenkins_candidates
  • validate_box_jenkins_candidates
  • select_box_jenkins_model
  • forecast_box_jenkins

Each result retains the metadata needed to audit the step rather than exposing only the final coefficients.

Diagnostics

  • durbin_watson_test
  • breusch_godfrey, breusch_godfrey_raw
  • ljung_box, box_pierce
  • jarque_bera, normality_ks
  • breusch_pagan, arch_test
  • mean_zero_test, variance_ratio_test
  • residual_correlogram, residual_diagnostics
  • roots_report, redundancy_check

Forecasting

  • naive_forecast, drift_forecast
  • prediction_interval
  • metrics
  • restore_differences

State-space

  • LinearStateSpace
  • kalman_filter, local_level_filter
  • kalman_smoother
  • kalman_forecast
  • estimate_local_level
  • kalman_innovation_diagnostics
  • state_space_adequacy
  • run_local_level_workflow

Plots and reports

  • plot_series
  • plot_correlogram, plot_eviews_correlogram
  • plot_decomposition
  • plot_forecast
  • result objects with summary(), table() and interpret() where provided.

For the conceptual explanation, start with the Time Series User Guide. For exact stochastic-process object signatures, use the existing class pages in docs/api/.