Mapping-patients

Geocoding and Mapping patients for IMB

This private repository provides code for geocoding and mapping patients for IMB.

Input files

This project is based on 3 input files that must be in the DATA/INPUT/ folder:

Demographic information: a set of addresses and personal patient information format (demog.xlsx). Only one column. Each address is a row.
Primary Visit: one row per patient, observations made during the first visit (baseline)
Subsequent Visit: several row per patient: one per subsequent visit

Files can be linked using the patient IDs. Every ~month, these files are updated on the Drupal server. Running the pipeline re-build the according Dataviz.

What the pipeline does

Step 1: Data wrangling: Read data, Clean it, Compute additional variables. Merge files together. Find the most recent observation in the Subsequent visit file for each patient.
Step 2: Geocoding: Extract the addresses that haven’t been geocoded yet. Geocode them. Use the opencage R library and update the list of geocoded addresses.
Step 3: Webpage input creation: Build Json object containing all the necessary data. These object are stored in .js files that will be read by the HTML final output
Output: one HTML file called index.html is available. It is the final output of this pipeline.

map

How to use the pipeline

Start by cloning this whole repository.
Update the input files of the DATA/INPUT folder with new data if needed
Run the pipeline to create the output. Must be done in the PIPELINE folder

./script_update_maps.sh

Open index.html to check if the webpage works properly

Requirements & limitations

A few R libraries need to be installed for the pipeline to work properly

library(dplyr)
library(lubridate)
library(readxl)
library(tidyr)
library(opencage)
library(jsonlite)

Note that the geocoding step is limited to 2500 calls / day.

Note on MND

Motor Neurone Disease (MND) is a disease that causes the death of neurons controling voluntaring muscles. It is also known under the name Amyotrophic lateral sclerosis (ALS).