progress on #82

R/make_mask_lookup.R

@@ -0,0 +1,29 @@
+#' .. content for \description{} (no empty lines) ..
+#'
+#' .. content for \details{} ..
+#'
+#' @title
+#' @param model_data
+#' @return
+#' @author Emma Mendelsohn
+#' @export
+make_mask_lookup <- function(model_dates_selected, rsa_polygon) {
+
+  masked_dates_90_days_lookup <- map_dfr(model_dates_selected, function(date){
+    diffs <- model_dates_selected - date
+    tibble(date = date, mask = list(model_dates_selected[diffs > 0 & diffs <= 90]))
+  })
+
+  masked_shapes_adjacent_lookup <- map_dfr(1:nrow(rsa_polygon), function(i){
+    select_shape <- rsa_polygon[i,]
+    touches <- st_touches(select_shape, rsa_polygon)
+    touches_shapes <- rsa_polygon[unlist(touches),]
+    tibble(shape = select_shape$shapeName, 
+           mask = list(touches_shapes$shapeName))
+  })
+
+  list(masked_dates_90_days_lookup = masked_dates_90_days_lookup,
+       masked_shapes_adjacent_lookup = masked_shapes_adjacent_lookup
+       )
+
+}

_targets.R

@@ -537,17 +537,19 @@ model_targets <- tar_plan(
 
   # Splitting --------------------------------------------------
   # Initial train and test (ie holdout)
-  tar_target(model_data_split, initial_split(model_data, prop = 0.8)), # pick random days left out of the training
+  tar_target(model_data_split, initial_split(model_data, prop = 0.8)), # pick random days and shapes to be withheld from training
   tar_target(training_data, training(model_data_split)),
   tar_target(holdout_data, testing(model_data_split)),
 
   # CV splits
   # Mask from the training set the three months following the holdout dates for the given district and the surrounding districts. 
   # Should this be on the whole training set, or just the analysis portion of the training set?
-  # In other words, it doesn't have to be masked from the model's assessments in the CV routine, we can use that data to assess performance
-  tar_target(masked_data, get_masks_for_training(holdout_data)),
+  # In other words, it doesn't have to be masked from the model's assessments in the CV routine, we can still use that data to assess performance
+  tar_target(mask_lookup, make_mask_lookup(model_dates_selected, rsa_polygon)),
+  tar_target(holdout_data_masks, holdout_data, mask_lookup), # TODO use mask_lookup against holdout_data to determine which dates/shapes need to be excluded from training
   tar_target(training_splits, vfold_cv(training_data)), # subsplit training analysis/assessment
-  tar_target(training_splits_masked, 1), # subsplit training analysis/assessment
+  tar_target(training_splits_masked, training_splits, holdout_data_masks), # TODO remove data from holdout_data_masks in analysis splits
+  # TODO in addition to the above step, we need to get assessment data masks from the splits, and mask these from each assessment split
 
   # Define formula and model
   tar_target(model_workflow, build_workflow(training_data)),