robyn_outputs: Evaluate Models and Output Results into Local Files

Description

Pack robyn_plots(), robyn_csv(), and robyn_clusters() outcomes on robyn_run() results. When UI=TRUE, enriched OutputModels results with additional plots and objects.

Create a plot to visualize the convergence for each of the datasets when running robyn_run(), especially useful for when using ts_validation. As a reference, the closer the test and validation convergence points are, the better, given the time-series wasn't overfitted.

Usage

robyn_outputs(
  InputCollect,
  OutputModels,
  pareto_fronts = "auto",
  calibration_constraint = 0.1,
  plot_folder = NULL,
  plot_folder_sub = NULL,
  plot_pareto = TRUE,
  csv_out = "pareto",
  clusters = TRUE,
  select_model = "clusters",
  ui = FALSE,
  export = TRUE,
  all_sol_json = FALSE,
  quiet = FALSE,
  refresh = FALSE,
  ...
)
# S3 method for robyn_outputs
print(x, ...)
robyn_csv(
  InputCollect,
  OutputCollect,
  csv_out = NULL,
  export = TRUE,
  calibrated = FALSE
)
robyn_plots(
  InputCollect,
  OutputCollect,
  export = TRUE,
  plot_folder = OutputCollect$plot_folder,
  ...
)
robyn_onepagers(
  InputCollect,
  OutputCollect,
  select_model = NULL,
  quiet = FALSE,
  export = TRUE,
  plot_folder = OutputCollect$plot_folder,
  baseline_level = 0,
  ...
)
ts_validation(OutputModels, quiet = FALSE, ...)
decomp_plot(
  InputCollect,
  OutputCollect,
  solID = NULL,
  exclude = NULL,
  baseline_level = 0
)

Value

(Invisible) list. Class: robyn_outputs. Contains processed results based on robyn_run() results.

Invisible NULL.

Invisible list with ggplot plots.

Invisible list with patchwork plot(s).

Invisible list with ggplot plots.

Arguments

InputCollect, OutputModels: robyn_inputs() and robyn_run() outcomes.
pareto_fronts: Integer. Number of Pareto fronts for the output. pareto_fronts = 1 returns the best models trading off NRMSE & DECOMP.RSSD. Increase pareto_fronts to get more model choices. pareto_fronts = "auto" selects the min fronts that include at least 100 candidates. To customize this threshold, set value with min_candidates.
calibration_constraint: Numeric. Default to 0.1 and allows 0.01-0.1. When calibrating, 0.1 means top 10 selection. Lower calibration_constraint increases calibration accuracy.
plot_folder: Character. Path for saving plots and files. Default to robyn_object and saves plot in the same directory as robyn_object.
plot_folder_sub: Character. Sub path for saving plots. Will overwrite the default path with timestamp or, for refresh and allocator, simply overwrite files.
plot_pareto: Boolean. Set to FALSE to deactivate plotting and saving model one-pagers. Used when testing models.
csv_out: Character. Accepts "pareto" or "all". Default to "pareto". Set to "all" will output all iterations as csv. Set NULL to skip exports into CSVs.
clusters: Boolean. Apply robyn_clusters() to output models?
select_model: Character vector. Which models (by solID) do you wish to plot the one-pagers and export? Default will take top robyn_clusters() results.
ui: Boolean. Save additional outputs for UI usage. List outcome.
export: Boolean. Export outcomes into local files?
all_sol_json: Logical. Add all pareto solutions to json export?
quiet: Boolean. Keep messages off?
refresh: Boolean. Refresh mode
...: Additional parameters passed to robyn_clusters()
x: robyn_outputs() output.
OutputCollect: robyn_run(..., export = FALSE) output.
calibrated: Logical
baseline_level: Integer, from 0 to 5. Aggregate baseline variables, depending on the level of aggregation you need. Default is 0 for no aggregation. 1 for Intercept only. 2 adding trend. 3 adding all prophet decomposition variables. 4. Adding contextual variables. 5 Adding organic variables. Results will be reflected on the waterfall chart.
solID: Character vector. Model IDs to plot.
exclude: Character vector. Manually exclude variables from plot.