From d89d6e316f65451d0a1ca457984f84ada8ee8aca Mon Sep 17 00:00:00 2001
From: Milan Wiedemann <milan.wiedemann@gmail.com>
Date: Mon, 2 Dec 2024 09:07:08 +0000
Subject: [PATCH] Refactor data loading and tidying into separate scripts

This will improve readability and modularity and make it easier for us to make further changes later on. It's only a start, there's much more we could do.
---
 lib/functions/load_opensafely_outputs.R | 88 +++++++++++++++++++++++++
 lib/functions/load_validation_data.R    | 42 ++++++++++++
 lib/functions/plot_measures.R           |  6 ++
 lib/functions/tidy_measures.R           | 35 ++++++++++
 4 files changed, 171 insertions(+)
 create mode 100644 lib/functions/load_opensafely_outputs.R
 create mode 100644 lib/functions/load_validation_data.R

diff --git a/lib/functions/load_opensafely_outputs.R b/lib/functions/load_opensafely_outputs.R
new file mode 100644
index 0000000..46204bd
--- /dev/null
+++ b/lib/functions/load_opensafely_outputs.R
@@ -0,0 +1,88 @@
+# Load data based on execution environment
+if (Sys.getenv("OPENSAFELY_BACKEND") != "") {
+  # Load data from generate_pf_measures action
+  df_measures <- readr::read_csv(
+    here("output", "measures", "pf_codes_conditions_measures.csv")
+  )
+  df_descriptive_stats <- read_csv(
+    here("output", "measures", "pf_descriptive_stats_measures.csv")
+  )
+  df_pfmed <- read_csv(
+    here("output", "measures", "pf_medications_measures.csv")
+  )
+  df_condition_provider <- read_csv(
+    here("output", "measures", "pf_condition_provider_measures.csv")
+  )
+} else {
+  # Load data from released_output directory
+  df_measures <- readr::read_csv(
+    here("released_output", "measures", "pf_codes_conditions_measures.csv")
+  )
+  df_descriptive_stats <- read_csv(
+    here("released_output", "measures", "pf_descriptive_stats_measures.csv")
+  )
+  df_pfmed <- read_csv(
+    here("released_output", "measures", "pf_medications_measures.csv")
+  )
+  df_condition_provider <- read_csv(
+    here("released_output", "measures", "pf_condition_provider_measures.csv")
+  )
+}
+
+df_measures <- tidy_measures(
+  data = df_measures,
+  pf_measures_name_dict = pf_measures_name_dict,
+  pf_measures_name_mapping = pf_measures_name_mapping,
+  pf_measures_groupby_dict = pf_measures_groupby_dict
+)
+
+df_measures$ethnicity <- factor(
+  df_measures$ethnicity,
+  levels = c(
+    "White",
+    "Mixed",
+    "Asian or Asian British",
+    "Black or Black British",
+    "Chinese or Other Ethnic Groups",
+    "Missing"
+  ),
+  ordered = TRUE
+)
+
+df_measures$age_band <- factor(
+  df_measures$age_band,
+  levels = c(
+    "0-19",
+    "20-39",
+    "40-59",
+    "60-79",
+    "80+",
+    "Missing"
+  ),
+  ordered = TRUE
+)
+
+df_measures$region <- factor(
+  df_measures$region,
+  levels = c(
+    "East",
+    "East Midlands",
+    "London",
+    "North East",
+    "North West",
+    "South East",
+    "South West",
+    "West Midlands",
+    "Yorkshire and The Humber",
+    "Missing"
+  ),
+  ordered = TRUE
+)
+
+df_measures <- df_measures %>%
+  mutate(sex = factor(sex,
+    levels = c("female", "male"),
+    labels = c("Female", "Male")
+  ))
+
+df_measures$age_band[is.na(df_measures$age_band)] <- "Missing"
\ No newline at end of file
diff --git a/lib/functions/load_validation_data.R b/lib/functions/load_validation_data.R
new file mode 100644
index 0000000..539dc41
--- /dev/null
+++ b/lib/functions/load_validation_data.R
@@ -0,0 +1,42 @@
+library(readr)
+library(tidyr)
+library(dplyr)
+library(here)
+
+df_bsa_consultation_validation <- read_csv(
+  here("lib", "validation", "data", "pf_consultation_validation_data.csv")
+) %>%
+  rename(count_100pct = count) |>
+  mutate(count_40pct = round(as.numeric(count_100pct * .4), digits = 0)) %>%
+  mutate(source = "nhs_bsa") |>
+  pivot_longer(
+    cols = c(count_100pct, count_40pct),
+    names_to = "count_method",
+    values_to = "count"
+  )
+
+df_bsa_consultation_validation <- df_bsa_consultation_validation %>%
+  mutate(consultation_type = factor(consultation_type,
+    levels = c(
+      "sinusitis",
+      "infected_insect_bites",
+      "uncomplicated_uti",
+      "acute_otitis_media",
+      "acute_sore_throat",
+      "shingles",
+      "impetigo"
+    ),
+    labels = c(
+      "Acute Sinusitis",
+      "Infected Insect Bite",
+      "UTI",
+      "Acute Otitis Media",
+      "Acute Pharyngitis",
+      "Herpes Zoster",
+      "Impetigo"
+    )
+  ))
+
+df_bsa_medication_validation <- read_csv(
+  here("lib", "validation", "data", "pf_medication_validation_data.csv")
+)
diff --git a/lib/functions/plot_measures.R b/lib/functions/plot_measures.R
index 9525dfa..045ebcb 100644
--- a/lib/functions/plot_measures.R
+++ b/lib/functions/plot_measures.R
@@ -97,3 +97,9 @@ plot_measures <- function(
 
   plot_tmp
 }
+
+# Colour palettes
+gradient_palette <- c("#001F4D", "#0056B3", "#007BFF", "#66B3E2", "#A4D8E1", "grey")
+region_palette <- c("red", "navy", "#018701", "#ffa600ca", "purple", "brown", "#f4a5b2", "cyan", "green", "grey")
+ethnicity_palette <- c("#42db0188", "#0056B3", "#ff0000c2", "#a52a2a5a", "purple", "grey")
+sex_palette <- c("red", "blue")
\ No newline at end of file
diff --git a/lib/functions/tidy_measures.R b/lib/functions/tidy_measures.R
index 2689ac4..44ab4e6 100644
--- a/lib/functions/tidy_measures.R
+++ b/lib/functions/tidy_measures.R
@@ -1,3 +1,38 @@
+# Define dictionaries with tidy names and mappings for measures
+pf_measures_name_dict <- list(
+  consultation_service = "Consultation Service",
+  pharmacy_first_service = "Pharmacy First Consultation",
+  combined_pf_service = "Pharmacy First Consultations (Combined)",
+  acute_otitis_media = "Acute Otitis Media",
+  herpes_zoster = "Herpes Zoster",
+  acute_sinusitis = "Acute Sinusitis",
+  impetigo = "Impetigo",
+  infected_insect_bite = "Infected Insect Bite",
+  acute_pharyngitis = "Acute Pharyngitis",
+  uncomplicated_urinary_tract_infection = "UTI"
+)
+
+pf_measures_name_mapping <- list(
+  consultation_service = "clinical_service",
+  pharmacy_first_service = "clinical_service",
+  combined_pf_service = "pharmacy_first_services",
+  acute_otitis_media = "clinical_condition",
+  herpes_zoster = "clinical_condition",
+  acute_sinusitis = "clinical_condition",
+  impetigo = "clinical_condition",
+  infected_insect_bite = "clinical_condition",
+  acute_pharyngitis = "clinical_condition",
+  uncomplicated_urinary_tract_infection = "clinical_condition"
+)
+
+pf_measures_groupby_dict <- list(
+  age_band = "Age band",
+  sex = "Sex",
+  imd = "IMD",
+  region = "Region",
+  ethnicity = "Ethnicity"
+)
+
 #' Tidy measures data
 #'
 #' Creates a tidier dataframe of measures data.