plot_hierarchical_forecasts: Create stacked plot of hierarchical forecasts

In ing-bank/tsforecast: Time Series Forecasting Pipeline

Description

plot_hierarchical_forecasts is a function that creates a plot which looks at how the forecasts of the chosen hierarchical group compares to it's children, if applicable.

Usage

plot_hierarchical_forecasts(
  main_forecasting_table,
  fc_model = c("consistent", "bottom_up"),
  hierarchical_cols = "",
  grouping = "",
  demo_mode = FALSE
)

Arguments

main_forecasting_table	A tibble containing a unique value for ts_split_date but not for the grouping column. It is assumed that this table is created using the create_main_forecasting_table function and which has been extended with the fc_models and fc_errors columns using the add_fc_models_to_main_forecasting_table and add_hierarchical_fc_models_to_main_forecasting_table functions.
fc_model	A string that can be either "consistent" for consistent hierarchical forecast models or "bottom_up" for bottom-up hierarchical forecast models.
hierarchical_cols	A string indicating which grouping columns are the hierarchical ones to be used in the plot.
grouping	A string indicating which grouping is the top group of the stacked plot.
demo_mode	Boolean, which is to be set to TRUE if any potentially sensitive figures should be hidden from the audience for demo purposes, or set to FALSE if all figures can safely be displayed.

Value

A plotly object displaying a stacked plot, if the selected grouping is a hierarchical parent.

Examples

dummy_hierarchical_gasprice %>%
      tstools::initialize_ts_forecast_data(
      date_col = "year_month",
      col_of_interest = "gasprice",
      group_cols = "currency",
      hierarchical_cols = c("location", "oil_company")
   ) %>%
   dplyr::filter(period >= as.Date("2004-11-30")) %>%
   create_main_forecasting_table() %>%
   add_fc_models_to_main_forecasting_table(
      fc_methods = c("basic", "linear")
   ) %>%
   add_hierarchical_fc_models_to_main_forecasting_table() %>%
   dplyr::filter(ts_split_date == 200611) %>%
   plot_hierarchical_forecasts(
      fc_model = "consistent",
      hierarchical_cols = "location",
      grouping = "location = USA   &   oil_company = CompanyC   &   currency = EUR"
   )

ing-bank/tsforecast documentation built on Sept. 18, 2020, 9:40 a.m.