Data From Part 1
ihsg_tbl <- read_rds("ihsg_tbl")
jkse_current_tbl <- read_rds("jkse_current_tbl")
ihsg_future_tbl <- read_rds("ihsg_future_tbl")
splits <- read_rds("splits")
regression_model_tbl <- read_rds("regression_model_tbl")
machine_learning_model_tbl <- read_rds("machine_learning_model_tbl")
boosted_model_tbl <- read_rds("boosted_model_tbl")
all_model_tbl <- read_rds("all_model_tbl")
Resampling (Time Series Split Cross Validation)
resamples_tscv_ihsg <- time_series_cv(
data = training(splits),
cumulative = TRUE,
initial = "5 year", # train
assess = "2 year", # test
skip = "2 year",
slice_limit = 5,
date_var = date
)
resamples_tscv_ihsg %>% write_rds("resamples_tscv_ihsg")
resamples_tscv_ihsg <- read_rds("resamples_tscv_ihsg")
resamples_tscv_ihsg %>%
tk_time_series_cv_plan() %>%
plot_time_series_cv_plan(date, value)
Regression Model
regression_model_resample_fitted <- regression_model_tbl %>%
modeltime_fit_resamples(resamples_tscv_ihsg,
control = control_resamples(verbose = FALSE))
regression_model_resample_fitted %>% write_rds("regression_model_resample_fitted")
regression_model_resample_fitted <- read_rds("regression_model_resample_fitted")
regression_model_resample_fitted %>%
plot_modeltime_resamples()
## Warning: There were 2 warnings in `dplyr::summarise()`.
## The first warning was:
## ℹ In argument: `.estimate = metric_fn(truth = .value, estimate = .pred, na_rm =
## na_rm)`.
## ℹ In group 8: `.resample_id = "Slice2"`, `.model_desc = "5_BATS(0.089, {0,0},
## 0.972, -)"`.
## Caused by warning:
## ! A correlation computation is required, but `estimate` is constant and has 0 standard deviation, resulting in a divide by 0 error. `NA` will be returned.
## ℹ Run ]8;;ide:run:dplyr::last_dplyr_warnings()dplyr::last_dplyr_warnings()]8;; to see the 1 remaining warning.
regression_model_resample_fitted %>%
modeltime_resample_accuracy(summary_fns = mean) %>%
table_modeltime_accuracy(.interactive = FALSE)
## Warning: There were 2 warnings in `dplyr::summarise()`.
## The first warning was:
## ℹ In argument: `.estimate = metric_fn(truth = value, estimate = .pred, na_rm =
## na_rm)`.
## ℹ In group 17: `.model_id = 5`, `.model_desc = "BATS(0.089, {0,0}, 0.972, -)"`,
## `.resample_id = "Slice2"`, `.type = "Resamples"`.
## Caused by warning:
## ! A correlation computation is required, but `estimate` is constant and has 0 standard deviation, resulting in a divide by 0 error. `NA` will be returned.
## ℹ Run ]8;;ide:run:dplyr::last_dplyr_warnings()dplyr::last_dplyr_warnings()]8;; to see the 1 remaining warning.
## Warning: There was 1 warning in `dplyr::summarise()`.
## ℹ In argument: `dplyr::across(.fns = summary_fns, ...)`.
## Caused by warning:
## ! Using `across()` without supplying `.cols` was deprecated in dplyr 1.1.0.
## ℹ Please supply `.cols` instead.
| Accuracy Table |
| .model_id |
.model_desc |
.type |
n |
mae |
mape |
mase |
smape |
rmse |
rsq |
| 1 |
ARIMA(4,1,2)(0,0,2)[5] WITH DRIFT |
Resamples |
5 |
948.04 |
32.81 |
31.49 |
33.85 |
1054.44 |
0.36 |
| 3 |
PROPHET |
Resamples |
5 |
653.45 |
23.75 |
22.23 |
21.65 |
743.01 |
0.50 |
| 4 |
PROPHET |
Resamples |
5 |
653.33 |
23.74 |
22.23 |
21.63 |
742.88 |
0.50 |
| 5 |
BATS(0.089, {0,0}, 0.972, -) |
Resamples |
5 |
590.13 |
21.01 |
20.34 |
22.46 |
678.09 |
NA |
Machine Learning Model
machine_learning_model_resample_fitted <- machine_learning_model_tbl %>%
modeltime_fit_resamples(resamples_tscv_ihsg,
control = control_resamples(verbose = FALSE))
machine_learning_model_resample_fitted %>% write_rds("machine_learning_model_resample_fitted")
machine_learning_model_resample_fitted <- read_rds("machine_learning_model_resample_fitted")
machine_learning_model_resample_fitted %>%
plot_modeltime_resamples()
machine_learning_model_resample_fitted %>%
modeltime_resample_accuracy(summary_fns = mean) %>%
table_modeltime_accuracy()
Boosted Model
boosted_model_resample_fitted <- boosted_model_tbl %>%
modeltime_fit_resamples(resamples_tscv_ihsg,
control = control_resamples(verbose = FALSE))
boosted_model_resample_fitted %>% write_rds("boosted_model_resample_fitted")
boosted_model_resample_fitted <- read_rds("boosted_model_resample_fitted")
boosted_model_resample_fitted %>%
plot_modeltime_resamples() # Slice 2
boosted_model_resample_fitted %>%
modeltime_resample_accuracy(summary_fns = mean) %>%
table_modeltime_accuracy()
boosted_model_tbl %>%
modeltime_accuracy(new_data = testing(splits)) %>%
# modeltime_resample_accuracy(summary_fns = mean) %>%
table_modeltime_accuracy()
Accuracy Checking
all_model_accuracy_tbl <- all_model_tbl %>%
modeltime_calibrate(new_data = testing(splits)) %>%
modeltime_accuracy()
all_model_resample_fitted_accuracy_tbl <- all_model_tbl_resample_fitted %>%
modeltime_calibrate(new_data = testing(splits)) %>%
modeltime_accuracy()
all_model_accuracy_tbl %>% write_rds("all_model_accuracy_tbl")
all_model_accuracy_tbl <- read_rds("all_model_accuracy_tbl")
all_model_resample_fitted_accuracy_tbl %>% write_rds("all_model_resample_fitted_accuracy_tbl")
all_model_resample_fitted_accuracy_tbl <- read_rds("all_model_resample_fitted_accuracy_tbl")
all_model_accuracy_tbl %>%
table_modeltime_accuracy(defaultPageSize = 30, bordered = TRUE, resizable = TRUE)
all_model_resample_fitted_accuracy_tbl %>%
table_modeltime_accuracy(defaultPageSize = 30, bordered = TRUE, resizable = TRUE)
Selected Model based on RMSE
all_model_accuracy_tbl %>%
table_modeltime_accuracy(defaultPageSize = 30, bordered = TRUE, resizable = TRUE)
# Model ID : Propet #4 (533.23), Prophet #3 (533.51), Prophet Boost #15 (674.89), ARIMA Boost #16 (839.28)
all_model_resample_fitted_accuracy_tbl %>%
table_modeltime_accuracy(defaultPageSize = 30, bordered = TRUE, resizable = TRUE)
# Model ID : Propet #4 (533.23), Prophet #3 (533.51), Prophet Boost #15 (674.89), ARIMA Boost #16 (839.28)