This is a prototype!

Coming soon:

• How to train brand models
• How to use the matching assessment and recalculate the models with the positive lists
• Explanations on the filter options
• Swagger API documentation
• Activate GPU acceleration for the matching and model creation process

Run as Docker Container (Flask + Gunicorn)

1. Requirements

You need docker and docker-compose installed!

2. Install PostgreSQL-Service with docker

• Set your desired postgres_user | postgres_pass and postgres_dbname --> Remember to adjust those changes in the ofs_settings_docker.yml if you don't use the standard values:

`docker run --name=ofs-postgis -d -e POSTGRES_USER=gis_admin -e POSTGRES_PASS=admin -e POSTGRES_DBNAME=gis -e ALLOW_IP_RANGE=0.0.0.0/0 -p 5434:5432 kartoza/postgis:9.6-2.4`

3. Prepare the current brand matching models

• Download the models: Download link is coming soon.
• Prepare the models: Create a folder called ann_models in the ofs root folder and copy the models in it.

4. Start OpenFuelService

Run ofs from the root folder:

docker-compose up -d

5. Configure OpenFuelService

• Create the db:

docker exec -it container_name /ofs_venv/bin/python manage.py create_db

• Import the data:

docker exec -it container_name /ofs_venv/bin/python manage.py import_data

• Match the data (This will take some time):

docker exec -it container_name /ofs_venv/bin/python manage.py match_data

• If done, ofs should be ready to test. Head to the end of the file to find some test queries.

Manual Setup (Ubuntu and Arch)

Ubuntu Preparations

1. Requirements

• You need at least Ubuntu 18.04 with python 3.6!
• A freshly created virtual environment (recommended!)

2. Install the required packages

sudo apt update
sudo apt install python3-distutils  
sudo apt install python3-dev

Arch Preparations

1. Requirements

• A freshly created virtual environment (recommended!)
• Install the beautiful yay aur manager

2. Install required packages

sudo pacman -S base-devel  
sudo pacman -S postgresql-libs  
sudo pacman -S libyaml  
sudo pacman -S cython  
sudo pacman -S cython2  
sudo pacman -S geos  
yay -S python-distutils  
yay -S python2-distutils  

General steps

3. Activate your venv

4. Install requirements

• Install the requirements.txt from the ofs base folder

pip install -r requirements.txt   

5. Install PostgreSQL-Service with docker

• Add your desired postgres_user | postgres_pass and postgres_dbname --> Remember to adjust those changes in the ofs_settings.yml if you don't use the standard values:

docker run --name=ofs-postgis -d -e POSTGRES_USER=gis_admin -e POSTGRES_PASS=admin -e POSTGRES_DBNAME=gis -e ALLOW_IP_RANGE=0.0.0.0/0 -p 5434:5432 kartoza/postgis:9.6-2.4

6. Prepare the matching models

• Download the models: Download link is coming soon.
• Prepare the models: Copy the models in the openfuelservice/server/files/models/ folder.

7. Create the database

• Run the creation process from the root folder of ofs:

python manage.py create-db   

• Couldn't drop user gis_admin and Couldn't create user gis_admin can be ignored!

8. Import the data

python manage.py import-data

8. Run the server

python manage.py run  

Test queries

• Please adjust the curl requests if you changed anything from the standard settings e.g. port and ip from the docker container etc.

[GET] Standard data queries:

• Get supported categories:

curl -X GET \
  'http://127.0.0.1:5000/fuel?request=categories' \
  -H 'cache-control: no-cache'

• Get currently supported brands:

curl -X GET \
  'http://127.0.0.1:5000/fuel?request=brands&source=cfd' \
  -H 'cache-control: no-cache'

• Get currently supported cars for VW:

curl -X GET \
  'http://127.0.0.1:5000/fuel?request=cars&brand=VW&source=cfd' \
  -H 'cache-control: no-cache'

[POST] Route Calculation examples:

• Query for all categories:

curl -X POST \
'http://127.0.0.1:5000/fuel?request=route' \
-H 'Content-Type: application/json' \
-H 'cache-control: no-cache' \
-d '{
  "request": "route",
  "geometry": {
    "geojson": {
      "coordinates":  [
      [8.679111522296585, 49.40613104364613],
      [8.679119403627244, 49.40613014781563]
      ],
      "type": "LineString"
    },
    "filters":{
      "data_source": "cfd",
      "fuel_type": "gasoline",
      "vehicle_type": "car",
      "driving_speed": "150",
      "vehicle_categories": ["all"],
      "request_id": "test123"
    }
  }
}'

• Query only category b and c and set the tank_size and an individual fuel_consumption to override ofs values:

curl -X POST \
'http://127.0.0.1:5000/fuel?request=route' \
-H 'Content-Type: application/json' \
-H 'cache-control: no-cache' \
-d '{
  "request": "route",
  "geometry": {
    "geojson": {
      "coordinates":  [
      [8.679111522296585, 49.40613104364613],
      [8.679119403627244, 49.40613014781563]
      ],
      "type": "LineString"
    },
    "filters":{
      "data_source": "cfd",
      "fuel_type": "gasoline",
      "vehicle_type": "car",
      "driving_speed": "150",
      "vehicle_categories": ["b", "c"],
      "tank_sizes": {"a":"20"},
      "fuel_consumptions": {"a": "4.5"},
      "request_id": "test123"
    }
  }
}'

• Query a specific car from the database. You need to query a hash from the database to paste it in the body (See the example above: Get currently supported cars for VW ):

curl -X POST \
  'http://127.0.0.1:5000/fuel?request=route' \
  -H 'Content-Type: application/json' \
  -H 'cache-control: no-cache' \
  -d '{
  "request": "route",
  "geometry": {
    "geojson": {
      "coordinates": [
                    [
                        10.502782,
                        51.181212
                    ],
                    [
                        10.50239,
                        51.1812
                    ],
                    [
                        10.501769,
                        51.181171
                    ]
                ],
                "type": "LineString"
    },
     "filters":{
	  "data_source": "cfd",
      "fuel_type": "gasoline",
      "vehicle_type": "car",
      "driving_speed": "60",	
      "cfd_ids": ["{enter your car id here without the {}}"],
      "request_id": "test123"
    }
  }
}'