readd(data_REE) %>%
plot_time_series(date, value, .interactive = interactive)
readd(splits_REE) %>%
tk_time_series_cv_plan() %>%
plot_time_series_cv_plan(date, value, .interactive = FALSE)
readd(models_tbl_REE)
#> # Modeltime Table
#> # A tibble: 4 x 3
#> .model_id .model .model_desc
#> <int> <list> <chr>
#> 1 1 <fit[+]> ARIMA(0,1,0) WITH DRIFT
#> 2 2 <fit[+]> ARIMA(0,1,0) WITH DRIFT W/ XGBOOST ERRORS
#> 3 3 <fit[+]> ETS(M,AD,M)
#> 4 4 <fit[+]> PROPHET
readd(calibration_tbl_REE)
#> # Modeltime Table
#> # A tibble: 4 x 5
#> .model_id .model .model_desc .type .calibration_data
#> <int> <list> <chr> <chr> <list>
#> 1 1 <fit[+]> ARIMA(0,1,0) WITH DRIFT Test <tibble [59 x 4]>
#> 2 2 <fit[+]> ARIMA(0,1,0) WITH DRIFT W/ XGBOOST ERRORS Test <tibble [59 x 4]>
#> 3 3 <fit[+]> ETS(M,AD,M) Test <tibble [59 x 4]>
#> 4 4 <fit[+]> PROPHET Test <tibble [59 x 4]>
readd(forecast_tbl_REE) %>%
plot_modeltime_forecast(.legend_max_width = 25,
.interactive = interactive)
#> Warning in max(ids, na.rm = TRUE): no non-missing arguments to max; returning -Inf
readd(accuracy_tbl_REE)$`_data`
#> # A tibble: 4 x 9
#> .model_id .model_desc .type mae mape mase smape rmse rsq
#> <int> <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 1 ARIMA(0,1,0) WITH DRIFT Test 3.98 5.74 4.21 5.58 4.61 0.44
#> 2 2 ARIMA(0,1,0) WITH DRIFT W/ XGBOOST ERRORS Test 4.12 5.94 4.35 5.75 4.78 0.44
#> 3 3 ETS(M,AD,M) Test 3.06 4.39 3.24 4.32 3.6 0.51
#> 4 4 PROPHET Test 4.1 5.71 4.33 5.86 4.92 0.44
readd(two_week_fc_REE)
#> # A tibble: 5 x 6
#> .ticker .index .value .low .high .model_desc
#> <chr> <date> <dbl> <dbl> <dbl> <chr>
#> 1 REE 2022-01-03 69.3 63.3 75.2 ETS(M,AD,M)
#> 2 REE 2022-01-04 69.2 63.2 75.1 ETS(M,AD,M)
#> 3 REE 2022-01-05 69.1 63.1 75.0 ETS(M,AD,M)
#> 4 REE 2022-01-06 69.2 63.3 75.2 ETS(M,AD,M)
#> 5 REE 2022-01-07 69.2 63.2 75.1 ETS(M,AD,M)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.