Merge pull request #120 from microsoft/users/mitokic/cran-0.2.2

Users/mitokic/cran 0.2.2

Author: AKannanMSFT

CRAN-SUBMISSION

@@ -1,3 +1,3 @@
-Version: 0.2.1
-Date: 2022-11-14 16:00:19 UTC
-SHA: 210b2963207f3dae698abbef41ab9d016ffab32f
+Version: 0.2.2
+Date: 2023-02-12 00:21:23 UTC
+SHA: 8ae9d89cd57696d8f528305401f874f7890a338e

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: finnts
 Title: Microsoft Finance Time Series Forecasting Framework
-Version: 0.2.1.9000
+Version: 0.2.2
 Authors@R: 
     c(person(given = "Mike",
            family = "Tokic",

NEWS.md

@@ -1,4 +1,4 @@
-# finnts 0.2.1.9000 DEVELOPMENT VERSION
+# finnts 0.2.2
 
 ## Bug Fixes
 

tests/testthat/test-forecast_time_series.R

@@ -13,7 +13,7 @@ check_exist <-function(to_check,ret){
 forecast_horizon <- 3
 target_variable <- "value"
 combo_variables <- c("id")
-models_to_run <- c("meanf", "snaive")
+models_to_run <- "snaive"
 inp_data <- modeltime::m750 %>% 
   dplyr::rename(Date = date) %>% 
   dplyr::mutate(id = as.character(id)) %>%
@@ -27,13 +27,13 @@ finn_forecast <- forecast_time_series(
   date_type = dt_type, 
   forecast_horizon = forecast_horizon, 
   run_model_parallel = FALSE,
-  back_test_scenarios = 6,
+  back_test_scenarios = 3,
   models_to_run = models_to_run, 
   run_global_models = FALSE, 
   run_ensemble_models = FALSE)
 
 final_fcst <- finn_forecast$final_fcst %>%
-  mutate(Date=as.Date(Date))
+  dplyr::mutate(Date=as.Date(Date))
 back_test_data <- finn_forecast$back_test_data
 back_test_best_MAPE <- finn_forecast$back_test_best_MAPE
 
@@ -135,7 +135,9 @@ finn_forecast <- forecast_time_series(
   run_model_parallel = FALSE,
   back_test_scenarios = 3,
   models_to_run = models_to_run, 
-  run_global_models = FALSE)
+  recipes_to_run = "R1",
+  run_global_models = FALSE, 
+  run_ensemble_models = FALSE)
 
 final_fcst <- finn_forecast$final_fcst %>%
   mutate(Date=as.Date(Date))
@@ -217,7 +219,9 @@ inp_data <- hts::infantgts %>%
     dplyr::mutate(Date = as.Date(paste0(index, "-07-01"))) %>%
     dplyr::select(-c("Total", "Sex/female", "Sex/male", dplyr::contains("State/"))) %>%
     tidyr::pivot_longer(-c("index", "Date"), names_to = "id") %>%
-    tidyr::separate(id, sep = " ", into = c("State", "Sex"), remove = FALSE)
+    tidyr::separate(id, sep = " ", into = c("State", "Sex"), remove = FALSE) %>%
+    dplyr::filter(Date >= "1985-07-01", 
+                  State %in% c("NSW", "VIC"))
 
 inp_data_combos <- inp_data %>%
   dplyr::mutate(Combo = paste0(State, Sex))
@@ -234,7 +238,9 @@ finn_forecast <- forecast_time_series(
   run_model_parallel = FALSE,
   back_test_scenarios = 3,
   models_to_run = models_to_run, 
-  run_global_models = FALSE)
+  recipes_to_run = "R1",
+  run_global_models = FALSE, 
+  run_ensemble_models = FALSE)
 
 final_fcst <- finn_forecast$final_fcst %>%
   mutate(Date=as.Date(Date))