PaperTrade is a trading app (fake money) that uses the Mboum API for market information (real data).
Data is persisted in a MongoDB cluster. The GUI is made with CustomTKinter. Multithreading was used to have extra information logged to the terminal while fetching data (great for debugging), and to ensure long calls don't block the GUI.
First, install some dependencies:
pip install customtkinter # The GUI
pip install dotenv # Environment Vars
pip install pymongo # MongoDB API for python
pip install requests # HTTP requests library
pip install mplcursors # To hover over the graph
Then create a .env file at the root. Set MONGO_CONNECTION_STRING to a connection string of a blank MongoDB cluster, and MBOUM_API_KEY to a MBOUM API key.
Finally, run python main.py in a terminal.
- Add an options tab.
Mboum does have an options REST endpoint, but integrating it with this app would be slightly different. To ensure strike prices are detected accordingly and expiration dates aren't missed, it would be important to have the option to keep this app running as a background process/thread. Historical data calls (as what's being done right now) would require several polls on startup to ensure we didn't miss reaching a strike price.
Currently, multithreading is used only for the API calls and MongoDB fetches, but we can set the main thread to be a daemon, polling Mboum routinely instead of polling several times on launch. To avoid polling entirely, we could alternatively use some websocket interface (not Mboum).
- Dockerize
It would be nice to keep the same database across all users of this app (which would imply restructuring the DB slightly to account for multiple users), but this would be easier and safer after dockerizing.