This is just a mushroom heatmap built on top of vk geotagged photos. The repository contains both client and backend code.
Client code is located under /docs directory, so that the frontend from the web can access it: geomushroom.
Backend consists of scrapper, which retrives data from vk and should be run beforehand, and serving application, which just gives clients pieces of information retrieved by scrapper beforehand.
- Create virtualenv with python3
- Install requirements:
pip install -r requirements.txt
- Install redis for caching and run it
- Follow the link from get_token.txt and copy result token from the address bar.
- Copy file config_template.yaml and paste obtained token under
vk_token
key - Prepare data sinks:
- If you want to store results in firebase, create firebase db.
- For postgres sink, to initialize database you can use
initdb.sql
from the root directory:psql -a -f initdb.sql
- Note however, that postgres sink is not yet ready to be used on client side :)
- Setup data sinks configuration in previously copied
config.yaml
- Perform data retrivement from vk by calling
./scrapper.py
- To build client js code
- Install npm and necessary packages
- Run
tsc --outFile ../../docs/geomushroom.js geomushroom.ts
fromsrc/client
directory
- If you want to use own server, you need to change ip hardcoded in docs/geomushroom.js to the one belonging to your server.
- Open docs/geomushroom.html. This is the user interface to control scrapper.
P.S. detected mushrooms can be poisonous. Use at your own risk.
Result exampleScrapper image classification is backed by tfhub models