apy.py

import pandas as pd


def pnl_per_trade(entry_usd, exit_usd):
    """
    pnl denom in usd for each trade
    """
    pnl_usd = exit_usd - entry_usd
    return pnl_usd


def return_per_trade(entry_usd, exit_usd):
    """
    return for each trade
    """
    pnl_usd = pnl_per_trade(entry_usd, exit_usd)
    ret = pnl_usd / entry_usd
    return ret


def get_trades_per_user(user_id):
    """
    TODO: write this function to query users trades  by their ids in the form defined below
    """
    trades = {
        "id1": {
            "entry_timestamp": "2023-06-12 00:00:00",
            "entry_usd": 110,
            "exit_timestamp": "2023-06-12 12:00:00",
            "exit_usd": 120,
        },
        "id2": {
            "entry_timestamp": "2023-06-12 13:00:00",
            "entry_usd": 125,
            "exit_timestamp": "2023-06-12 23:00:00",
            "exit_usd": 120,
        },
    }

    return trades


def get_returns(trades: dict):
    rets = pd.Series()
    for trade_id in trades:
        entry_time = pd.to_datetime(trades[trade_id]["entry_timestamp"])
        entry_usd = trades[trade_id]["entry_usd"]
        exit_usd = trades[trade_id]["exit_usd"]
        ret = return_per_trade(entry_usd, exit_usd)
        rets.loc[entry_time] = ret

    return rets


def get_duration(start, end):
    dur = end - start
    dur_days = dur.days + dur.seconds / 3600 / 24
    return dur_days


def get_apy(user_id):
    trades = get_trades_per_user(user_id)
    rets = get_returns(trades)
    cumret = rets.sum()
    duration = get_duration(rets.index[0], rets.index[-1])
    if duration != 0:
        cumret_annual = cumret * 365 / duration
    else:
        cumret_annual = 0
    return cumret, cumret_annual


if __name__ == "__main__":
    print(get_apy(""))