From 8050822a63fde8116181fa0817880d42e51492c6 Mon Sep 17 00:00:00 2001
From: Dan Snow <31494343+dfsnow@users.noreply.github.com>
Date: Fri, 20 Dec 2024 13:52:21 -0600
Subject: [PATCH] Fix missing `geography_id` value for triads (#299)

* Fix missing geography_id value for triads

* Fix styler issue
---
 pipeline/03-evaluate.R | 32 +++++++++++++++++++++-----------
 1 file changed, 21 insertions(+), 11 deletions(-)

diff --git a/pipeline/03-evaluate.R b/pipeline/03-evaluate.R
index 8c0480bf..7f72cb5c 100644
--- a/pipeline/03-evaluate.R
+++ b/pipeline/03-evaluate.R
@@ -61,6 +61,23 @@ assessment_data_pin <- read_parquet(paths$output$assessment_pin$local) %>%
     pred_pin_final_fmv_round, sale_ratio_study_price
   )
 
+# Helper function to return NA when sale sample size is too small
+gte_n <- \(n_sales, min_n, fn, na_type) {
+  if (sum(!is.na(n_sales)) >= min_n) {
+    return(fn)
+  } else {
+    return(na_type)
+  }
+}
+
+# Helper function to add triad code as geography ID if it's not already present
+add_triad_code <- \(data) {
+  if (!"geography_id" %in% colnames(data)) {
+    data$geography_id <- data$triad_code
+  }
+  return(data)
+}
+
 
 
 
@@ -73,15 +90,6 @@ assessment_data_pin <- read_parquet(paths$output$assessment_pin$local) %>%
 gen_agg_stats <- function(data, truth, estimate, bldg_sqft,
                           rsn_col, rsf_col, triad, geography,
                           class, col_dict, min_n) {
-  # Helper function to return NA when sale sample size is too small
-  gte_n <- \(n_sales, min_n, fn, na_type) {
-    if (sum(!is.na(n_sales)) >= min_n) {
-      return(fn)
-    } else {
-      return(na_type)
-    }
-  }
-
   # List of summary stat/performance functions applied within summarize() below
   # Each function is listed on the right while the name of the function is on
   # the left
@@ -239,7 +247,8 @@ gen_agg_stats <- function(data, truth, estimate, bldg_sqft,
     mutate(across(
       -(contains("_max") & contains("yoy")) & where(is.numeric),
       ~ replace(.x, !is.finite(.x), NA)
-    ))
+    )) %>%
+    add_triad_code()
 }
 
 
@@ -299,7 +308,8 @@ gen_agg_stats_quantile <- function(data, truth, estimate,
     mutate(across(
       -(contains("_max") & contains("yoy")) & where(is.numeric),
       ~ replace(.x, !is.finite(.x), NA)
-    ))
+    )) %>%
+    add_triad_code()
 }