Disclaimer: Do not use it at home if you don't know how to :)

You will lose money. Created only for educational purposes.

AV-Binance-Trading

Trading service for different investment strategies based on technical indicators and tweeter post analysis executed with binance API. Historic data is acquired from yahoo and binance streams.
There are different type of granularity: stream, minute, daily.
Some of the strategies need to be run in parallel with crypto streams to database.

Prerequisites:

• Python 3.9
• Poetry 1.7.1
• Docker 20.10.22
• Make 4.2.1

Quickstart

• make install Virtual environment is managed with poetry package manager. After executing poetry shell on host machine you have access to cli with commands to run different strategies.

• make up-frontend Frontend of this app is a streamlit dashboard with top trading currencies and other functionalities.

  • http://localhost:8501/home - dashboard

• make up Backend is FastAPI with endpoints for managing strategies on host for specified currencies. It supports interactive swagger menu.

  • http://localhost:8000/docs - endpoint documentation
  • http://localhost:8000/redoc - interactive endpoint documentation

• make test Tests can be run on live services on host or on mocked data acquired from .csv files depending on TEST_ON_LIVE_SERVICESvariable to choice.

Use cases:

• multiple users can use this app and log in/ sign in into dashboard
• user can enter dashboard and view top trading currencies
• user can start trading with different strategies
• user can view profit reports from different strategies

make help:

Please use make target where target is one of:
clean               clean up temp and trash files
format              format code
help                display help message
install             install dependencies
lint                run static code checkers
test                run pytest
up-front            run dashboard
up                  run API

Example scripts need to be run with python starting in examples/ directory!

If make test won't run try to make clean, there are problems with cache.

Main bot is trading currencies defined in position_checks table, it can be reloaded within reloads_position_checks() from file defined tests/test_data_position_checks_basic.csv.

Project structure:

• core - main directory of a project
  • streams.py - methods used for starting streaming data to database
  • utils.py - utilities for project management
  • strategies.py - main file containing all investment strategies, they can be run using features scripts
  • models.py - class models
  • converters.py - data converters, adding, naming, converting columns in dataframes
  • constants.py - constants
  • binance.py - integration with binance API, creating orders, get historical data
  • backtesting.py - backtesting performance of strategies on historical data
• features - directory with python scripts used to run an investment strategy
  • streamlit - live dashboard with currency price
• graphics - graphics services used to visualize data and plot currency value, histograms of tweet analysis
• predictions
  • data_extractors.py - different historical data sources for prediction models
  • predict_models.py - fbprophet model for currency forecasting
• twitter
  • tweety - integration with tweety api, get tweets
• top_worst_crypto_view.py - top performing currencies view

Environment

Create .env file:

Variable	Description	Value
api_key	Binance TestNet api key
api_secret	Binance TestNet api secret
API_VALID_KEY	Binance API valid key
API_VALID_SECRET	Binance API secret key
consumer_key	Tweety API consumer key
consumer_secret	Tweety API consumer secret key
bearer_token	Tweety API bearer token
access_token	Tweety API access token
access_token_secret	Tweety API secret access token
TEST_ON_LIVE_SERVICES	Should tests run with real services?	True/False

For testing purposes you can obtain TestNet tokens from: https://testnet.binance.vision/
Click log in with GitHub and then generate token: https://testnet.binance.vision/key/generate

Poetry

Top used commands for poetry package manager:
poetry shell - enter shell
exit - exit the shell
poetry add pytest --dev - add requirement
poetry update - update version
poetry install - install poetry
poetry run binance_trading - main command
poetry env use /home/lorbi/.cache/pypoetry/virtualenvs/av-crypto-trading-84RDaCA4-py3.9/bin/python - use new python
poetry lock --no-update - update lock file after chaning pyproject.toml poetry self update - update the version of poetry
my interpreter: /home/lorbi/.cache/pypoetry/virtualenvs/av-crypto-trading-q9kMY0JF-py3.9/bin/python

Technologies:

• python
• pandas
• decorators
• streamlit
• websocket
• ta
• numpy
• fbprohet
• nltk
• spacy
• re
• plotly
• sqlalchemy
• fastapi
• pytest
• docker
• Github Actions
• click
• poetry
• prospector
• Makefile

Common issues:

• error: command '/usr/bin/x86_64-linux-gnu-gcc' failed with exit code 1

sudo apt-get install python3.9-dev build-essential

Git large files:

sudo apt-get install git-lfs
git lfs install
git lfs track "*.csv"
git lfs pull