VITObelgium / guppy

A Python web API for easy geospatial analysis of raster data

Quickstart

Assuming Docker and docker-compose are installed on your system, you can start the Guppy server with the following commands:

cd server
docker-compose up 

Then go access the demo UI https://github.com/VITObelgium/guppy/tree/develop?tab=readme-ov-file#access-the-ui or interact with the API at http://localhost:8080/guppy/docs

Installation

Instead of using the docker-compose file, you can also run the Guppy server locally. This allows you to run the server in a development environment and to make changes to the code.

Conda environment

Ensure miniconda is installed on your system. If not, download and install it from here.

conda env create --file environment.yml

Usage

Configuration

Configuration file: config.yml

Create a configuration file in the directory server named config.yml with the following content. Adjust for your specific setup. minimal yaml:

# ./server/config.yml
database:
  type: sqlite
  db: guppy.sqlite
deploy:
  path: guppy
  content: /content

full yaml:

database:
  type: sqlite
  db: guppy.sqlite
deploy:
  path: guppy
  content: /content
guppy:
  size_limit: 2510000000
geoserver:
  username: geoserver_account
  password: geoserver_password

if you prefer a postgresql database change the database section to:

database:
  host: example-host.vito.be
  user: johndoe
  passwd: top-secret-password
  db: guppy

Configuration via environment variables

Alternatively, you can set the configuration via environment variables. These variables will override the values in the configuration file.

The following environment variables are available:

Environment variable	Description
GUPPY_CONFIG_FILE	Path to the configuration file
GUPPY_DEPLOY_PATH	Path to deploy the Guppy server. default: guppy
GUPPY_DEPLOY_CONTENT	Path of the content directory. default: /content
GUPPY_SIZE_LIMIT	Maximum input area size in m² for stats/data calls
GUPPY_DATABASE_TYPE	Database type (sqlite, postgres) default: sqlite
GUPPY_DATABASE_HOST	Database host
GUPPY_DATABASE_DB	Database name OR path to Sqlite file.
GUPPY_DATABASE_USER	Database user
GUPPY_DATABASE_PASSWD	Database password
GUPPY_DATABASE_PASSWD_FILE	Path to a file containing the database password
GUPPY_GEOSERVER_USER	GeoServer account username
GUPPY_GEOSERVER_PASSWD	GeoServer account password
GUPPY_GEOSERVER_PASSWD_FILE	Path to a file containing the password for the GeoServer account

bold items are mandatory for correct working of the server.

Start server

# activate conda environment
conda activate guppy

# change directory to 'server'
cd server

# start server using defaults
PYTHONPATH=. python guppy/server.py

Your empty Guppy server is available at http://127.0.0.1:5000/guppy/docs

Run the server with reload option to automatically restart the server when changes are made to the code.

python -m uvicorn guppy.server:app --reload --port 5000 --host 0.0.0.0

Access the UI

The Guppy API is the core of the application, some UI is implemented on top of the API to illustrate the functionality.

The pages already provided as examples are:

Page	URL
Upload layer	http://localhost:5000/guppy/admin/upload/ui
Layer metadata	http://localhost:5000/guppy/admin/layers
Tile statistics	http://localhost:5000/guppy/admin/stats
Layers viewer	http://localhost:5000/guppy/admin/map

Access the API

Documentation for the API is available at http://127.0.0.1:5000/guppy/docs

Notice deploy_path is defined in the config.yml file. The default value is guppy.