Skip to content

Latest commit

 

History

History
144 lines (83 loc) · 3.29 KB

setup.md

File metadata and controls

144 lines (83 loc) · 3.29 KB
Raw

Setup

This doc guides you through the setup of the Safer Route application.

Mapillary API

Get your API key here: https://www.mapillary.com/developer/api-documentation?locale=de_DE

put it into a config/config.json file:

{
  "mapillary_api_key": "your key here"
}

Network graph via openrouteservice

Run openrouteservice locally to export the pedestrian graph


wget http://download.geofabrik.de/europe/germany/baden-wuerttemberg/karlsruhe-regbez-latest.osm.pbf -N -O data/geofabrik_karlsruhe-regbez-latest.osm.pbf

wget http://download.geofabrik.de/europe/germany/bayern/oberbayern-latest.osm.pbf -N -O data/geofabrik_oberbayern-latest.osm.pbf

wget http://download.geofabrik.de/north-america/us/district-of-columbia-latest.osm.pbf -N -O data/geofabrik_district-of-columbia-latest.osm.pbf

osmium extract -b 8.412699,49.409633,8.590426,49.591490 data/geofabrik_karlsruhe-regbez-latest.osm.pbf -o data/ma.pbf --overwrite

osmium extract -b 11.35963,48.06082,11.72441,48.25068 data/geofabrik_oberbayern-latest.osm.pbf -o data/munich.pbf --overwrite 

osmium extract -b -77.12119,38.79080,-76.90884,38.99678 data/geofabrik_district-of-columbia-latest.osm.pbf -o data/dc.pbf --overwrite

osmium merge data/ma.pbf data/munich.pbf data/dc.pbf -o data/osm_file.pbf --overwrite

Create the openrouteservice folders to not fiddle with permisions. Move the configuration (pedestrian only) to the corresponding folder.

mkdir -p docker/conf docker/elevation_cache docker/graphs docker/logs/ors docker/logs/tomcat docker/data

mv ors-config.json docker/conf/.

ORS_UID=${UID} ORS_GID=${GID} docker-compose -f ors-docker-compose.yml up

Run the script to export the graph for the cities

Rscript extract_network_grap.R

Data acquisition and processing

The rest of the data acquisiont happens in a different script.

Run the script src/data_acquisition_processing.R

A data/ folder will be created which contains all raw and processed data. The processed data is ready to be consumed by the shiny app:

  • munich.Rdata
  • dc.Rdata
  • ma.Rdata

Deploy webapp via rocker

Create directories to be mounted by the docker container

Folders

mkdir -p shiny_logs shiny_apps/data

  • shiny_logs store the logs of our shiny application

  • shiny_apps holds the app.R as well as the data it is consuming

Run the docker container

The docker container is based of rocker/geospatial:ubuntugis. We install some libraries used in acquiring, processing and visualizing the data. (L#4) And we install the shiny server. (L#10)

All of this happens in the Dockerfile.

RUID="$(id -u)" docker-compose up

Copy the shiny application and the data to the volume to be served.

Copy the App components

cp -r src/app shiny_apps/.
cp  src/app.R shiny_apps/.

Copy the .RData files

cp  data/*.RData shiny_apps/data/.

Your shiny_apps folder should look like this:

$ tree shiny_apps/
shiny_apps/
├── app
│   ├── server.R
│   └── ui.R
├── app.R
└── data
    ├── dc.RData
    ├── ma.RData
    └── munich.RData

2 directories, 6 files

The shiny server runs at port 3838. See the shiny application here:

http://localhost:3838