Merge pull request #211 from signaturescience/dev

v2.0.2 release
signaturescience · May 5, 2024 · cf16d54 · cf16d54
2 parents 75c5fbe + 44704f3
commit cf16d54
Show file tree

Hide file tree

Showing 4 changed files with 17 additions and 5 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: fiphde
 Title: Forecasting Influenza in Support of Public Health Decision Making
-Version: 2.0.1
+Version: 2.0.2
 Authors@R: 
     c(person(given = "VP",
              family = "Nagraj",

diff --git a/NEWS.md b/NEWS.md
@@ -1,3 +1,11 @@
+# fiphde 2.0.2
+
+## New features
+
+### NNETAR in ensemble
+
+As of this release, if the NNETAR model is specified in the `ts_fit_forecast()` function (i.e., "nnetar" is not `NULL` in the model list) then the model will be included in the ensemble. In previous releases, if the NNETAR method was used the results would be included in the returned forecast and fit objects but would not be ensembled with the other components. Note that currently *all* of the possible models (ARIMA, ETS, and NNETAR) must be specified to use the NNETAR in the ensemble.
+
 # fiphde 2.0.1
 
 ## Bug fixes

diff --git a/R/fiphde-package.R b/R/fiphde-package.R
@@ -132,5 +132,6 @@ if(getRversion() >= "2.15.1")  utils::globalVariables(c(".",
                                                         "reference_date",
                                                         "target_name",
                                                         "target_prob",
+                                                        "nnetar",
                                                         "."))
 
diff --git a/R/forecast.R b/R/forecast.R
@@ -122,10 +122,13 @@ ts_fit_forecast <- function(prepped_tsibble,
   # e.g. if an ARIMA model doesn't fit for one location but the ETS does, you still want the ETS model for that location.
   tsfit <- purrr::reduce(tsfit, dplyr::inner_join, by="location")
 
-  # Ensemble the ARIMA and ETS models
-  # Hard-coded for now - can always ensemble other models if/when we add them.
-  # fixme: this could be more flexible
-  if (ensemble & !is.null(models$arima) & !is.null(models$ets)) {
+  # Ensemble the models
+  # TODO: make this more flexible and not hard-coded in the future?
+  if (ensemble & !is.null(models$arima) & !is.null(models$ets) & !is.null(models$nnetar)) {
+    tsfit <-
+      tsfit %>%
+      dplyr::mutate(ensemble=(arima+ets+nnetar)/3)
+  } else if (ensemble & !is.null(models$arima) & !is.null(models$ets)) {
     tsfit <-
       tsfit %>%
       dplyr::mutate(ensemble=(arima+ets)/2)