The map is work in progress
The map itself is visible at: https://johnardavies.github.io/City_map/#
1. The general food outlet information This is obtained by geocoding the list of Liverpool food outlets on the Food Standards Agency website The original xml is Liverpool_food_outlets.xml which is in the Data folder inside the map_data_processsing folder processed usimg the the Liverpool_map.ipynb notebook. All the original data for this and other layers is in this folder.
For the purposes of the map food outlets that are wholesale or not normal consumer retail are present in the Food_outlets_geo.geojson, but are filtered out in the map's javascript. These being outlets of the type:
'Mobile caterer','Distributors/Transporters', 'Manufacturers/packers', 'Hospitals/Childcare/Caring Premises','School/college/university','Other catering premises','Hotel/bed & breakfast/guest house'Supermarkets are included as a separate layer as Supermarkets_geo.geojson. This layer is created with the Liverpool_map.ipynb notebook
2. The affordable food initatives (Food pantries, Free food initatives, Community cafes information \ This has been obtained from the Liverpool good food plan team, it was originally collected to feed into the Feeding Liverpool map\ The original excel file is Affordable_food_inititives_Liverpool.xlsx which is processed with the Liverpool_map.ipynb notebook to produce the Affordable_food_initatives.geojson
3. The emergency food initatives (Emergency food parcel providers and Emergency meal providers) \ This has been obtained from Liverpool good food plan team. The original file is emergency_food_providers.xlsx which is processed with the Liverpool_map.ipynb notebook to produce the Emergency_food_suppliers.geojson which is read in by the map
4. The walking distance layer \ This is created by calling the MapboxGL isochrone API to calculate a 10 minutes walking isochrone. These are produced by using jQuery AJAX in the map's javascript to make API calls for respectively the coordinates of the food support initatives (Emergency food and affordable food initatives) and for the supermarkets, creating two sets of isochrone layers that are made visible to clicking the buttons. In principle these calls can be made separately, stored and then imported as static layers rather than using API calls.
5. The reception class obesity layer at the ward level \ This data has been obtained from Liverpool council. It is included in the Liverpool_wards.geojson as it is at ward level. There is other information at ward level and which is imported in the javascript, but it is the obesity layer which is shown in the map.
6. The e-food deserts index layer \ This is an index measuring the degree to which areas have characteristics associated with food deserts. The index is produced by CDRC and was obtained from their website It is included in the Liverpool_lsoa.geojson
7. The layers showing areas scoring in the bottom 20% and more than 10 minutes from either an emergency or an affordable food initative (These are shown by clicking the top checkbox) \These are produced using the Further_analysis.ipynb notebook. There is a geojson for each layer above the checkbox (With the exception of the travel time layers).
The layers are income.geojson, population.geojson, education.geojson, morbidity.geojson, obesity.geojson, crowding.geojson, score.geojson. The last two being for the household overcrowding and e-food index respectively.
8. The layers showing the levels of Food bank and pantry usage \ These are produced in a separate analysis which is private owing to the data that it includes. The geojson that produces these at 3-4 digit postcode level is Foodbank_pantry_use.geojson
9. The layers showing the connection between foodbank and pantry users and the centroids of the postcodes that make use of them (This is shown by clicking the bottom checkbox) \ These are created using a separate notebook as the data used to produce them is not public. The files that the map is reading are the Foodbank_network.geojson and the Pantry_network.geojson. The Food bank network has been filtered so that postcodes and foodbanks needed to be connected by at least 10 voucher uses to be included.
10. All other layer data \ This is taken from the 2019 index of multiple deprivation accessed on the ONS website on 19/05/21. The data is File 6: population denominators (saved in the Data folder of map_data_processing as File6-_IoD2019_Population_Denominators.xlsx') and File 8: underlying indicators (saved in the data folder as File8-_IoD2019_Underlying_Indicators.xlsx). These files are processed with Liverpool_map.ipynb to produce the Liverpool_lsoa.geojson which the map imports.
The processing of the data in the map is done using the notebooks which are run in a Docker container:
1. Liverpool_map.ipynb \ This generates the geojson files for the layers (isochrone layers and the two layers triggered by the checkboxes excepted) and the food outlet sites
2. Further_analysis.ipynb \ This generates geojson for areas that are in the lowest 20% of layers data and more than 10 mins from an affordable food outlet
3. The generation of the network layers that shows links between foodbanks and postcodes \ This is done using a third notebook. It is currently private due to this data not being public
The first two notebooks are run in sequence (Further_analysis.ipynb depends on the outputs of Liverpool_map.ipynb) within a container using the Dockerfile in the map_data_processing folder which also contains the data in the Data folder.
They process the data which is in the Data folder inside map_data_processing folder and then output the files which the map uses to the Out folder
To build the container image for the data processing after cloning the repo, tagging it as map_processing
$ docker build --tag map_processing map_data_processingTo run the container on port 8009 (mapped to the container port 8009)
$ docker run -p 8009:8009 --init map_processingTo download the processed files for the map from the Out folder in the container
$ docker cp container_id:/notebooks/Data/Out Destination_locationTo run the map locally
$ git clone https://github.com/johnardavies/City_mapand from inside the City_map folder in the same directory as the index file run the below, setting port_id to the port you want to see map on e.g. 8011
$ python -m http.server port_idThe map will then be visible on http://localhost:port_id e.g. http://localhost:8011