Merge pull request #88 from opensafely/github-codespaces

Add GitHub Codespaces

.devcontainer/README.md

@@ -0,0 +1,9 @@
+# Devcontainer Configuration
+
+This folder contains configuration files required for Codespaces/devcontainers.
+Please do not modify its contents unless you know what you are doing and there
+is a specific change to the devcontainer configuration you need to make.
+
+If you believe you have a broken configuration or are unsure if you have the
+latest version of the configuration, please follow the 
+[instructions to add codespaces to your project](https://docs.opensafely.org/getting-started/how-to/add-github-codespaces-to-your-project/).

.devcontainer/devcontainer.json

@@ -0,0 +1,57 @@
+// For format details, see https://aka.ms/devcontainer.json. For config options, see the
+// README at: https://github.com/devcontainers/templates/tree/main/src/python
+{
+    "name": "OpenSAFELY",
+    "image": "ghcr.io/opensafely-core/research-template:v0",
+    // Features to add to the dev container. More info: https://containers.dev/features.
+    "features": {
+        "ghcr.io/devcontainers/features/docker-in-docker:2": {}
+    },
+    "postCreateCommand": "/bin/bash /opt/devcontainer/postCreate.sh ${containerWorkspaceFolder}",
+    "postAttachCommand": "/bin/bash /opt/devcontainer/postAttach.sh",
+    "forwardPorts": [
+        8787
+    ],
+    "portsAttributes": {
+        "8787": {
+            "label": "RStudio IDE"
+        }
+    },
+    // Configure tool-specific properties.
+    "customizations": {
+        "codespaces": {
+            "repositories": {
+                "opensafely/server-instructions": {
+                    "permissions": {
+                        "contents": "read"
+                    }
+                }
+            }
+        },
+        "vscode": {
+            "extensions": [
+                "ms-python.python",
+                "ms-toolsai.jupyter",
+                "ms-toolsai.jupyter-renderers"
+            ],
+            "settings": {
+                "extensions.ignoreRecommendations": true,
+                "files.autoSave": "afterDelay",
+                "files.autoSaveDelay": 1000,
+                "git.autofetch": true,
+                "python.analysis.extraPaths": [
+                    ".devcontainer/ehrql-main/"
+                ],
+                "python.defaultInterpreterPath": "/opt/venv/bin/python",
+                "python.terminal.activateEnvInCurrentTerminal": true,
+                "python.terminal.activateEnvironment": true,
+                "window.autoDetectColorScheme": true
+            }
+        }
+    },
+    // Uncomment to connect as root instead. More info: https://aka.ms/dev-containers-non-root.
+    // "remoteUser": "root"
+    "remoteEnv": {
+        "MAX_WORKERS": "2"
+    }
+}

.gitignore

@@ -10,4 +10,6 @@ venv/
 */Rhistory
 .Rhistory
 */test.ipynb
+.devcontainer/ehrql-main/
+.DS_Store
 notebooks_jupyter/local_report_ethnicity.ipynb

ONS_check.R

@@ -0,0 +1,133 @@
+# Author: Colm D Andrews
+# Date:   14/07/2022
+#
+################################################################################
+library(tidyverse)
+library(scales)
+library(readr)
+library(glue)
+library(arrow)
+library(ggplot2)
+library(ggalluvial)
+
+### Point 1. Why is there weirdness with lower N in the grouped compared to overall England
+
+
+ethnicity <-
+  read_csv(here::here("output","released","simple_patient_counts_categories_5_group_registered.csv"),col_types =(cols())) %>% 
+  filter(group=="region") %>%
+  summarise(N=sum(Asian_ethnicity_new_5_filled))
+
+# Adding all regions gives 1706020 compared to 1708430 total ethnicities for Asian New codelist
+
+
+
+#######################################################
+
+ons_na_removed <-
+  read_csv(here::here("output", "released", "made_locally", "ethnic_group_2021_registered_with_2001_categories.csv")) %>%
+  mutate(
+    cohort = fct_relevel(cohort, "ONS", "new", "supplemented"),
+    Ethnic_Group = fct_relevel(
+      Ethnic_Group,
+      "Asian", "Black", "Mixed", "White", "Other"
+    )
+  )
+
+ons_england<-ons_na_removed %>%
+  filter(region!="England") %>%
+  group_by(Ethnic_Group,cohort) %>%
+  summarise(
+    region="England",
+    N_sumregion = sum(N),
+    Total_sumregion=sum(Total))
+
+ons_na_removed_compare <-ons_na_removed %>%
+  select("Ethnic_Group","region","cohort","N","Total")%>%
+  inner_join(ons_england) %>%
+  mutate(N_diff = N- N_sumregion,
+         total_diff = Total - Total_sumregion, 
+         perc_diff = N_diff / N *100)
+### some weirdness with lower N in the grouped* See point 1 above
+
+
+
+ons_england_remix <- ons_england %>%
+  summarise(
+            region,
+            Ethnic_Group,
+            Total = Total_sumregion,
+            N = N_sumregion,
+            group = 5,
+            cohort=cohort,
+            percentage = N/Total *100)
+
+#### Is only White overrepresented in England?
+ons_ethnicity_plot_na_diff <- ons_england_remix %>%
+  group_by(Ethnic_Group, region) %>%
+  arrange(cohort) %>%
+  mutate(diff = percentage - first(percentage)) %>%
+  select(region, Ethnic_Group, cohort, diff)
+
+
+## create difference in percentage between ONS and TPP (for plotting)
+ons_ethnicity_plot_na_diff <- ons_na_removed %>%
+  group_by(Ethnic_Group, region, group) %>%
+  arrange(cohort) %>%
+  mutate(diff = percentage - first(percentage)) %>%
+  filter(Ethnic_Group=="White",
+         cohort=="new"| cohort=="ONS") %>%
+  select(N, Total,cohort,region,percentage) %>%
+  pivot_wider(names_from = cohort,values_from = c(N,Total,percentage))
+
+# step 2: check England isa above ONS
+step2 <- ons_ethnicity_plot_na_diff %>% filter(region!="England") %>%
+  group_by(Ethnic_Group) %>%
+  summarise(perc_ONS = sum(N_ONS)/sum(Total_ONS),
+            perc_new = sum(N_new)/sum(Total_new))
+
+## Why is it above
+coverage<- ons_ethnicity_plot_na_diff %>%
+  filter(region!="England") %>%
+  mutate(
+    diff = percentage_new - percentage_ONS,
+    sign=case_when(
+              diff>0 ~"positive",
+              diff<0 ~"negative"),
+         coverage= Total_new / Total_ONS
+         )
+
+totals<-coverage %>% 
+  select(region,N_ONS,N_new,Total_ONS,Total_new,sign) %>%
+  group_by(sign) %>%
+  summarise(N_ONS = sum(N_ONS),
+            Total_ONS = sum(Total_ONS),
+            N_new= sum (N_new),
+            Total_new = sum(Total_new),
+            perc_ONS = N_ONS / Total_ONS*100,
+            perc_new = N_new / Total_new*100,
+            diff = perc_new - perc_ONS,
+            coverage= Total_new / Total_ONS) 
+
+# does this agree with step2?
+totals %>%
+  select(N_ONS,N_new,Total_ONS,Total_new) %>%
+  mutate(group=1) %>%
+  group_by(group) %>%
+  summarise(N_ONS = sum(N_ONS),
+            Total_ONS = sum(Total_ONS),
+            N_new= sum (N_new),
+            Total_new = sum(Total_new),
+            perc_ONS = N_ONS / Total_ONS*100,
+            perc_new = N_new / Total_new*100)
+
+step2
+
+### Does ONS discrepancy widen as -> 100%
+ons_discrepancy<- ons_ethnicity_plot_na_diff %>%
+  mutate(diff=percentage_new - percentage_ONS) 
+View(ons_discrepancy)
+
+
+
+