UltraArceus3/Spring-22-UConn-Hackathon

UConn Team 1's submission to the Spring 2022 Blockchain Hackathon

UConn Team 1's submission to the Spring 2022 Blockchain Hackathon.

Contributors:

Rany Kamel (UltraArceus3)

Nachiket Deo (Nachiket18)

Ya-sine Agrignan (Dev-10sky)

Thien-Bao (tbbui-732)

---

Technical Setup

1. Create a single github account with a public repository for each team and or individual if you are not with a team
2. Install git on your local machine
3. Install docker on your local machine (Optional)
4. Clone this repo git clone https://github.com/TradeBlock/tb-hackathon.git
5. install python version 3.10 on your local development environment.
6. run pip install -r requirements.txt in directory you cloned
7. create .env file # for AWS credentials

Optional Setup - Create Python Virtual Environment

1. Create the environment with python3.10 -m venv .venv
2. Activate the environment before step 8 in Technical Setup with source .venv/bin/activate
3. Continue with step 8 in Technical Setup

Problem Statement

Summary

Develop a trading strategy which will provide trading signals used to predict when to BUY or SELL a cryptocurrency pair based on previous data from different exchanges. Please review references for details around how bitcoin and ethereum internals work.

Constraints

A notional limit of 1 million USD will be used for limiting buys meaning the initial cash available to trade is limited to this amount Trades can only be a BUY or a SELL with a cryptocurrency amount SELL trades can only be done after BUYS (i.e. no short sell) No outside data will be allowed

Measures of Success

A stream of trades for a fixed period of time will be fed into a function which will return an enumeration of BUY, SELL, or NEUTRAL. The return value will then be used to simulate actual trades in the market and provide a monetary profit or loss. The ending notional value of your model based computation will determine your score. The score will determine the leaderboard and ultimately the prizes to be awarded.

Data

• Data will be available for bulk download in .csv files
• Data also can be easily downloaded into your program one row at a time in main.py
• Row data will be formatted as follows:
  • pair, price, amount, timestamp
  • pair: EXCHANGE-CRYPTO-usd where EXCHANGE is a specific exchange where the trade happened and CRYPTO is either bitcoin (xbt) or ethereum (eth)
  • price: price in dollars of the trade
  • amount: number of units which is either bitcoins or ethereum contracts
  • timestamp: is seconds since epoch and can be used in python via datetime.datetime.fromtimestamp(TIMESTAMP)
  • Sample rows are:
    • okfq-xbt-usd,14682.26,2,1514765115
    • okf1-xbt-usd,13793.65,2,1514765115
    • stmp-xbt-usd,13789.01,0.00152381,1514765115

Response Data Object Definitions

Fill

{
  “price”: decimal.Decimal,       # fill price
  “volume”: decimal.Decimal,      # fill volume
  “error_code”: str               # Optional. “rejected”
  “error_msg”: str                # Optional. description of error
  “unfilled”: dict[str, Decimal]  # unfilled portion of trades (xbt,eth)
}

Unfilled

{
  “xbt”: decimal.Decimal,         # unfilled xbt volume
  “eth”: decimal.Decimal,         # unfilled eth volume
}

Errors

When an error occurs with your attempted trade, you will receive an empty fill dictionary and the unfilled trades dictionary—this may be caused by:

• Insufficient funds
• Yielded a type other than Trade or None; or
• Invalid base currency in yielded Trade

References

• bitcoin
• ethereum