Multi-Model Time Series Forecasting Toolkit and Local App

---

Documentation for package ‘forecastsuite’ version 1.0.2

• DESCRIPTION file
• User guides, package vignettes and other documentation
• Package README

Help Pages

always_zero_dates	Find calendar dates that are always zero or missing
analyze_series	Analyze a series' shape for model recommendation
build_cv_folds	Build rolling-origin (walk-forward) cross-validation folds
collapse_to_period	Collapse multiple rows sharing a period into one
compose_date_parts	Assemble a date from separate Year/Quarter/Month/Day columns
compute_forecast_bias	Signed forecast error (bias)
compute_forecast_metrics	Compute MASE, sMAPE, and RMSE for a forecast
compute_group_value_map	Build a raw-value -> canonical-label map for a grouping column
compute_interval_coverage	Empirical prediction-interval coverage
compute_multi_window_metrics	Score a forecast over several evaluation windows at once
compute_residual_diagnostics	Residual diagnostics for a forecast
convert_months_to_horizon	Convert a forecast horizon in months to periods
count_duplicate_periods	Count how many rows share a period
decompose_series	Seasonal decomposition of a series
detect_anomalies	Flag unusual points in a series
detect_bias_drift	Check whether forecast bias is growing across the test window
ensemble_forecasts	Combine several models' forecasts into one
fixed_holiday_catalog	Built-in fixed-date holiday catalog
get_model	Look up a registered model
googlesheet_csv_url	Build a Google Sheets CSV export URL
group_correlation_matrix	Pairwise correlation between grouped series
holidays_with_data	Find declared holidays with non-zero data
holiday_contingency	Run both holiday consistency checks at once
holiday_limitation_note	Holiday-limitation note for a model
impute_anomalies	Impute detected anomalies before fitting
list_models	List every registered model
lstm_available	Check whether the LSTM model is available
plot_forecast_generic	Plot a single model's forecast
plot_model_comparison	Overlay several models' forecasts on one plot
process_uploaded_data	Turn an uploaded table into a ds/y forecasting series
prophet_available	Check whether the Prophet model is available
read_google_sheet	Read a link-shared Google Sheet as a data frame
recommend_model	Rank registered models for a series
reconcile_bottom_up	Bottom-up hierarchical reconciliation
reconcile_top_down	Top-down hierarchical reconciliation (historical proportions)
register_model	Register a forecasting model
run_app	Launch the bundled local forecastsuite Shiny app
run_backtest_leaderboard	Multi-model rolling-origin backtest leaderboard
safe_compute_metrics	Compute forecast metrics, guarding against a too-short test set
split_by_group	Split a ds/y/group tibble into one tibble per group
suggest_parameters	Suggest starting hyperparameters for one model
suggest_parameters_for	Suggest starting hyperparameters for several models at once
ts_frequency_for	Map an aggregation frequency to a ts() seasonal frequency