create lookup table for pulling dates from NDVI sources without expan…

…ding

R/create_ndvi_date_lookup.R

@@ -0,0 +1,47 @@
+#' .. content for \description{} (no empty lines) ..
+#'
+#' .. content for \details{} ..
+#'
+#' @title
+#' @param sentinel_ndvi_transformed
+#' @param sentinel_ndvi_transformed_directory
+#' @param modis_ndvi_transformed
+#' @param modis_ndvi_transformed_directory
+#' @return
+#' @author Emma Mendelsohn
+#' @export
+create_ndvi_date_lookup <- function(sentinel_ndvi_transformed,
+                                    sentinel_ndvi_transformed_directory,
+                                    modis_ndvi_transformed,
+                                    modis_ndvi_transformed_directory) {
+
+
+  sent_dataset <- open_dataset(sentinel_ndvi_transformed_directory) 
+  modi_dataset <- open_dataset(modis_ndvi_transformed_directory) 
+
+  # create lookup table so we know which rows to query, without doing an expansion on the actual data
+  sent_dates <- sent_dataset |> 
+    distinct(start_date, end_date) |> 
+    arrange(start_date) |> 
+    collect() |> 
+    mutate(lookup_dates = map2(start_date, end_date, ~seq(.x, .y-1, by = "1 day"))) |>  
+    mutate(satellite = "sentinel")
+
+  min_sent_start <- min(sent_dates$start_date) # this is when sentinel starts
+
+  modi_dates <- modi_dataset |> 
+    distinct(start_date, end_date) |> 
+    arrange(start_date) |> 
+    collect() |> 
+    mutate(lookup_dates = map2(start_date, end_date, ~seq(.x, .y-1, by = "1 day"))) |> 
+    mutate(satellite = "modis") |> 
+    mutate(lookup_dates = map(lookup_dates, ~na.omit(if_else(.>=min_sent_start, NA, .)))) |> 
+    filter(map_lgl(lookup_dates, ~length(.) > 0))
+
+  ndvi_dates <- bind_rows(modi_dates, sent_dates) |> 
+    mutate(lookup_day_of_year = map(lookup_dates, yday)) |> 
+    relocate(satellite)
+
+  return(ndvi_dates)
+
+}

_targets.R

@@ -297,7 +297,7 @@ data_targets <- tar_plan(
   tar_target(model_dates, set_model_dates(start_year = 2005, end_year = 2022, n_per_month = 2, lag_intervals, seed = 212)),
   tar_target(model_dates_selected, model_dates |> filter(select_date) |> pull(date)),
 
-  # weather data
+  # weather anomalies data
   tar_target(weather_historical_means_directory, 
              create_data_directory(directory_path = "data/weather_historical_means")),
 
@@ -334,6 +334,19 @@ data_targets <- tar_plan(
              pattern = weather_anomalies,
              cue = tar_cue("never")), # only run this if you need to upload new data  
 
+  # ndvi anomalies data
+  tar_target(ndvi_date_lookup, 
+             create_ndvi_date_lookup(sentinel_ndvi_transformed,
+                                     sentinel_ndvi_transformed_directory,
+                                     modis_ndvi_transformed,
+                                     modis_ndvi_transformed_directory)),
+
+
+  tar_target(ndvi_historical_means_directory, 
+             create_data_directory(directory_path = "data/ndvi_historical_means")),
+
+
+
 )
 
 # Model -----------------------------------------------------------