Summarise Forecasting Models Errror Metrics

Source: R/get_metrics.R

Summarise the out-of-sample stocks forecasts by asset and time periods.

get_metrics(.tbl, .group, .truth, .forecast)

Arguments

.tbl

A tibble object..
.group

The column in which the data should be grouped. This will often be a column with stock tickers or stocks names.
.truth

The truth results of a time series realization.
.forecast

The out-of-sample forecasts from the chosen model.

Value

A tibble with the following summary statistics by asset: mse, rmse, mae, mape and mase.

Examples

library(YahooTickers)
library(dplyr)
library(forecast)

get_tickers(dow) %>%
  slice(2:3) %>%
  get_stocks(., periodicity = "monthly") %>%
  get_returns(., tickers, log, TRUE, adjusted) %>%
  get_models(., tickers, adjusted, 100, 1, FALSE, auto.arima,
             seasonal = FALSE, stationary = TRUE) %>%
  get_forecasts(.) %>%
  get_metrics(., tickers, adjusted, point_forecast)
#> # A tibble: 2 × 6
#>   tickers     mse   rmse    mae  mape  mase
#>   <fct>     <dbl>  <dbl>  <dbl> <dbl> <dbl>
#> 1 AXP     0.00677 0.0823 0.0539 NaN   0.177
#> 2 AMGN    0.00447 0.0669 0.0524  94.4 0.161