freqtrade/tests/data/test_converter_public_trades.py

import logging
from pathlib import Path

import arrow
import numpy as np
import pandas as pd
import pytest
from pandas import DataFrame

from freqtrade.configuration import Configuration
from freqtrade.constants import DEFAULT_ORDERFLOW_COLUMNS
from freqtrade.data.converter import (populate_dataframe_with_trades, public_trades_to_dataframe,
                                      trades_to_volumeprofile_with_total_delta_bid_ask)
from freqtrade.enums import CandleType, MarginMode, TradingMode
from freqtrade.exchange.exchange import timeframe_to_minutes
from tests.conftest import get_mock_coro, get_patched_exchange, log_has, log_has_re, testdatadir


BIN_SIZE_SCALE = 0.5


def read_csv(filename, converter_columns: list = ['side', 'type']):
    return pd.read_csv(filename, skipinitialspace=True, infer_datetime_format=True, index_col=0,
                       parse_dates=True, converters={col: str.strip for col in converter_columns})


@pytest.fixture(scope="module")
def populate_dataframe_with_trades_dataframe():
    return pd.read_json('tests/testdata/populate_dataframe_with_trades_dataframe.json').copy()


@pytest.fixture(scope="module")
def populate_dataframe_with_trades_trades():
    # dataframe['date'] = pd.to_datetime(dataframe['date'], unit='ms', utc=True)
    return pd.read_feather('tests/testdata/populate_dataframe_with_trades_trades.feather').copy()


@pytest.fixture(scope="module")
def candles():
    return pd.read_json('tests/testdata/candles.json').copy()


@pytest.fixture(scope="module")
def trades():
    return pd.read_json('tests/testdata/trades.json').copy()


@pytest.fixture(scope="module")
def public_trades_list():
    return read_csv('tests/testdata/public_trades_list.csv').copy()


@pytest.fixture(scope="module")
def public_trades_list_simple():
    return read_csv('tests/testdata/public_trades_list_simple_example.csv').copy()


@pytest.fixture(scope="module")
def public_trades_list_simple_results():
    return read_csv('tests/testdata/public_trades_list_simple_results.csv').copy()


@pytest.fixture(scope="module")
def public_trades_list_simple_bidask():
    return read_csv('tests/testdata/public_trades_list_simple_bidask.csv').copy()


def conjuresetup():
    public_trades_list = public_trades_list()
    print(public_trades_list.columns.tolist())
    public_trades_list_simple = public_trades_list_simple()
    print(public_trades_list_simple.columns.tolist())
    print(public_trades_list_simple.loc[:, [
        'timestamp', 'id', 'price', 'side', 'amount']])
    public_trades_list_simple_results = public_trades_list_simple_results()
    print(public_trades_list_simple_results.columns.tolist())
    public_trades_list_simple_bidask = public_trades_list_simple_bidask()
    print(public_trades_list_simple_bidask.columns.tolist())
    print(public_trades_list_simple_bidask)
    print(public_trades_list_simple_results)
# conjuresetup()  # never called except in REPL
# /conjuresetup


def test_public_trades_columns_before_change(populate_dataframe_with_trades_dataframe, populate_dataframe_with_trades_trades):
    assert populate_dataframe_with_trades_dataframe.columns.tolist() == [
        'date', 'open', 'high', 'low', 'close', 'volume']
    assert populate_dataframe_with_trades_trades.columns.tolist() == [
        'timestamp', 'id', 'type', 'side', 'price',
        'amount', 'cost', 'date']


def load_entries_from_strategy(filepath: str,
                               *,
                               from_date: str = '',
                               configpath: str = "../volumio-strategy/user_data/config.json"):
    from freqtrade.data.dataprovider import DataProvider
    from freqtrade.resolvers import StrategyResolver
    """Load candle data from a file"""
    dataframe = pd.read_feather(f'{filepath}')
    # Convert date column to datetime
    # dataframe['date'] = pd.to_datetime(dataframe['date'])
    config = Configuration.from_files([configpath])

    # Define some constants
    config["timeframe"] = "1m"
    # Name of the strategy class
    config["strategy"] = "Volumio"
    # Pair to analyze - Only use one pair here
    pair = "BTC/USDT:USDT"

    # Load data
    candle_df = dataframe if (
        not from_date) else dataframe.loc[dataframe.date > from_date]
    # data from before we had imbalances
    if 'stacked_imbalances_bid' not in candle_df.columns:
        candle_df[['stacked_imbalances_bid',
                   'stacked_imbalances_ask']] = np.nan

    load_strategy = StrategyResolver.load_strategy(config)
    dataprovider = DataProvider(config, None, None)
    load_strategy.dp = dataprovider

    metadata = {'pair': pair}
    df = load_strategy.populate_entry_trend(
        load_strategy.populate_indicators(candle_df, metadata), metadata)
    plot_config = load_strategy.plot_config

    return (df, plot_config, pair)


def test_strategy_entries():
    from freqtrade.plot.plotting import generate_candlestick_graph

    """
    # for debug/adding new entries
    print(df.date.loc[df.is_enter_long>0])
    print(df.date.loc[df.is_enter_short>0])
    """
    df, plot_config, pair = load_entries_from_strategy(
        'tests/testdata/populate_indicators_dataframe-doubletop.feather')  # '2023-03-25 09:30:00+00:00')
    graph = generate_candlestick_graph(pair=pair,
                                       data=df,
                                       plot_config=plot_config,
                                       ).show()
    assert df['enter_long'].sum() > 0
    assert df['enter_short'].sum() > 0
    # 15:39 #4 "double top" entry
    assert df.loc[(865 < df.index) & (df.index <= 872)
                  ]['enter_short'].sum() > 0
    # 16:02 #5 strong hidden & regular divergences with reversal
    assert df.loc[(893 == df.index)]['enter_short'].sum() > 0
    # 17:08 #5 strong hidden & regular divergences with reversal
    assert df.loc[(957 <= df.index) & (df.index <= 962)
                  ]['enter_short'].sum() > 0

    df, plot_config, pair = load_entries_from_strategy(
        'tests/testdata/populate_indicators_dataframe-choppy.feather')  # '2023-03-25 09:30:00+00:00')
    graph = generate_candlestick_graph(pair=pair,
                                       data=df,
                                       plot_config=plot_config,
                                       ).show()
    assert df['enter_long'].sum() > 0
    assert df['enter_short'].sum() > 0
    # check good entries
    # 11:10
    assert df.iloc[565]['enter_short'] == 1
    # 06:16
    assert df.loc[(271 < df.index) & (df.index < 273)]['enter_long'].sum() > 0
    # 11:46 #3 below vwap
    assert df.loc[(600 < df.index) & (df.index < 610)]['enter_long'].sum() > 0
    # 13:07 #2 below 3rd vwap
    assert df.loc[(681 < df.index) & (df.index < 690)]['enter_long'].sum() > 0
    # 16:18
    assert df.loc[(873 < df.index) & (df.index <= 878)]['enter_long'].sum() > 0
    # 15:52
    assert df.loc[(845 < df.index) & (df.index <= 850)
                  ]['enter_short'].sum() > 0

    df, plot_config, pair = load_entries_from_strategy(
        'tests/testdata/populate_indicators_dataframe-downtrend.feather')  # '2023-03-25 09:30:00+00:00')
    graph = generate_candlestick_graph(pair=pair,
                                       data=df,
                                       plot_config=plot_config,
                                       ).show()
    assert df['enter_long'].sum() > 0
    assert df['enter_short'].sum() > 0
    # 07:50
    assert df.loc[(451 < df.index) & (df.index < 500)]['enter_long'].sum() > 0

    df, plot_config, pair = load_entries_from_strategy(
        'tests/testdata/populate_indicators_dataframe-uptrend.feather')  # '2023-03-25 09:30:00+00:00')
    graph = generate_candlestick_graph(pair=pair,
                                       data=df,
                                       plot_config=plot_config,
                                       ).show()
    assert df['enter_long'].sum() > 0
    assert df['enter_short'].sum() > 0
    # 09:23 - 11:55 long period of consolidation, then strong uptrend
    assert df.loc[(780 < df.index) & (df.index < 830)]['enter_long'].sum() > 0


def test_public_trades_mock_populate_dataframe_with_trades__check_orderflow(
        populate_dataframe_with_trades_dataframe,
        populate_dataframe_with_trades_trades):
    dataframe = populate_dataframe_with_trades_dataframe.copy()
    trades = populate_dataframe_with_trades_trades.copy()
    dataframe['date'] = pd.to_datetime(
        dataframe['date'], unit='ms').dt.tz_localize('UTC')
    dataframe = dataframe.copy().tail().reset_index(drop=True)
    config = Configuration.from_files(
        ["../volumio-strategy/user_data/config.json"])
    config['timeframe'] = '5m'
    config['orderflow']['scale'] = 0.005
    config['orderflow']['imbalance_volume'] = 0
    df = populate_dataframe_with_trades(config,
                                        dataframe, trades, pair='unitttest')
    results = df.iloc[0]
    t = results['trades']
    of = results['orderflow']
    assert 0 != len(results)  # 13 columns
    assert 4073 == len(t)

    # orderflow/cluster/footprint
    assert 506 == len(of)
    assert [39.0, 0.0, -22.598, 22.598, 0.0,
            22.598, 39.0] == of.iloc[0].values.tolist()
    assert [0.0, 4.0, 0.319, 0.0, 0.319, 0.319,
            4.0] == of.iloc[-1].values.tolist()
    of = df.iloc[-1]['orderflow']
    assert 434 == len(of)
    assert [18.0, 0.0, -3.367, 3.367, 0.0, 3.367,
            18.0] == of.iloc[0].values.tolist()
    assert [0.0, 3.0, 0.144, 0.0, 0.144, 0.144,
            3.0] == of.iloc[-1].values.tolist()

    assert -46.62299999999999 == results['delta']
    assert -97.12800000000034 == results['min_delta']
    assert 0.088 == results['max_delta']
    assert np.isnan(results['stacked_imbalances_bid'])
    assert 24219.7 == results['stacked_imbalances_ask']

    results = df.iloc[-3]
    assert 143.56099999999998 == results['delta']
    assert 0.0 == results['min_delta']
    assert 146.74999999999997 == results['max_delta']
    assert 24233.9 == results['stacked_imbalances_bid']
    assert np.isnan(results['stacked_imbalances_ask'])

    results = df.iloc[-1]
    assert 95.00900000000013 == results['delta']
    assert -8.579999999999998 == results['min_delta']
    assert 107.73599999999985 == results['max_delta']
    assert np.isnan(results['stacked_imbalances_bid'])
    assert np.isnan(results['stacked_imbalances_ask'])


def test_public_trades_trades_mock_populate_dataframe_with_trades__check_trades(
        populate_dataframe_with_trades_dataframe, populate_dataframe_with_trades_trades):
    dataframe = populate_dataframe_with_trades_dataframe.copy()
    trades = populate_dataframe_with_trades_trades.copy()

    # slice of unnecessary trades
    dataframe['date'] = pd.to_datetime(
        dataframe['date'], unit='ms').dt.tz_localize('UTC')
    # dataframe = dataframe.copy().reset_index(drop=True)
    dataframe = dataframe.copy().tail().reset_index(drop=True)
    trades = trades.copy().loc[trades.date >= dataframe.date[0]]
    trades.reset_index(inplace=True, drop=True)

    assert trades['id'][0] == '1637515870'

    config = {
        'timeframe': '5m',
        'orderflow': {'scale': 0.5, 'imbalance_volume': 0, 'imbalance_ratio': 300, 'stacked_imbalance_range': 3}
    }
    df = populate_dataframe_with_trades(config,
                                        dataframe, trades, pair='unitttest')
    result = df.iloc[0]
    assert result.index.values.tolist() == ['date', 'open', 'high', 'low', 'close', 'volume', 'trades', 'orderflow',
                                            'bid', 'ask', 'delta', 'min_delta', 'max_delta', 'total_trades', 'stacked_imbalances_bid', 'stacked_imbalances_ask']

    assert -46.62299999999999 == result['delta']
    assert 521.726 == result['bid']
    assert 475.103 == result['ask']

    assert 4073 == len(result.trades)
    t = result['trades'].iloc[0]
    assert trades['id'][0] == t["id"]
    assert int(trades['timestamp'][0]) == int(t['timestamp'])
    assert 'buy' == t['side']
    assert '1637515870' == t['id']
    assert 24229.1 == t['price']


def test_public_trades_cached_grouped_trades_pair(
        populate_dataframe_with_trades_dataframe, populate_dataframe_with_trades_trades):
    import time

    # slice of unnecessary trades
    dataframe_before = populate_dataframe_with_trades_dataframe.copy().head(
        20).reset_index(drop=True)
    dataframe_before['date'] = pd.to_datetime(
        dataframe_before['date'], unit='ms').dt.tz_localize('UTC')

    dataframe_next = populate_dataframe_with_trades_dataframe.copy().head(
        25).reset_index(drop=True)
    dataframe_next = dataframe_next.tail(24).reset_index(drop=True)
    dataframe_next['date'] = pd.to_datetime(
        dataframe_next['date'], unit='ms').dt.tz_localize('UTC')

    trades = populate_dataframe_with_trades_trades.copy()
    trades = trades.loc[trades.date >= dataframe_before.date[0]]
    trades = trades.loc[trades.date <= dataframe_before.iloc[-1].date]
    trades.reset_index(inplace=True, drop=True)
    start_time_before = time.time()
    config = {
        'timeframe': '5m',
        'orderflow': {'scale': 0.5, 'imbalance_volume': 0, 'imbalance_ratio': 300, 'stacked_imbalance_range': 3}
    }
    df = populate_dataframe_with_trades(config,
                                        dataframe_before, trades, pair='unitttest')
    end_time_before = time.time() - start_time_before
    # TODO: assert trades and delta received

    trades = populate_dataframe_with_trades_trades.copy()
    trades = trades.loc[trades.date >= dataframe_next.date[0]]
    trades = trades.loc[trades.date <= dataframe_next.iloc[-1].date]
    trades.reset_index(inplace=True, drop=True)
    start_time_next = time.time()
    df = populate_dataframe_with_trades(config,
                                        dataframe_next, trades,  pair='unitttest')
    end_time_next = time.time() - start_time_next
    # TODO: assert trades and delta received
    assert end_time_next < end_time_before


def test_public_trades_put_volume_profile_into_ohlcv_candles(public_trades_list_simple, candles):
    df = public_trades_to_dataframe(
        public_trades_list_simple, '1m', 'doesntmatter', fill_missing=False, drop_incomplete=False)
    df = trades_to_volumeprofile_with_total_delta_bid_ask(
        df, scale=BIN_SIZE_SCALE)
    candles['vp'] = np.nan
    candles.loc[candles.index == 1, ['vp']] = candles.loc[candles.index == 1, [
        'vp']].applymap(lambda x: pd.DataFrame(df.to_dict()))
    assert 0.14 == candles['vp'][1].values.tolist()[1][2]  # delta
    assert 0.14 == candles['vp'][1]['delta'].iat[1]


def test_public_trades_binned_big_sample_list(public_trades_list):
    BIN_SIZE_SCALE = 0.05
    trades = public_trades_to_dataframe(
        public_trades_list, '1m', 'doesntmatter',
        fill_missing=False, drop_incomplete=False)
    df = trades_to_volumeprofile_with_total_delta_bid_ask(
        trades, scale=BIN_SIZE_SCALE)
    assert df.columns.tolist() == ['bid', 'ask', 'delta',
                                   'bid_amount', 'ask_amount',
                                   'total_volume', 'total_trades']
    assert 23 == len(df)
    assert df.index[0] < df.index[1] < df.index[2]
    assert df.index[0] + BIN_SIZE_SCALE == df.index[1]
    assert (trades['price'].min() -
            BIN_SIZE_SCALE) < df.index[0] < trades['price'].max()
    assert (df.index[0] + BIN_SIZE_SCALE) >= df.index[1]
    assert (trades['price'].max() -
            BIN_SIZE_SCALE) < df.index[-1] < trades['price'].max()

    assert 32 == df['bid'].iat[0]  # bid
    assert 197.512 == df['bid_amount'].iat[0]  # bid
    assert 88.98 == df['ask_amount'].iat[0]  # ask
    assert 26 == df['ask'].iat[0]  # ask
    assert -108.53200000000001 == df['delta'].iat[0]  # delta
    assert 3 == df['bid'].iat[-1]  # bid
    assert 50.659 == df['bid_amount'].iat[-1]  # bid
    assert 108.21 == df['ask_amount'].iat[-1]  # ask
    assert 44 == df['ask'].iat[-1]  # ask
    assert 57.551 == df['delta'].iat[-1]  # delta

    BIN_SIZE_SCALE = 1
    trades = public_trades_to_dataframe(
        public_trades_list, '1m', 'doesntmatter',
        fill_missing=False, drop_incomplete=False)
    df = trades_to_volumeprofile_with_total_delta_bid_ask(
        trades, scale=BIN_SIZE_SCALE)
    assert 2 == len(df)
    assert df.index[0] < df.index[1]
    assert (trades['price'].min() -
            BIN_SIZE_SCALE) < df.index[0] < trades['price'].max()
    assert (df.index[0] + BIN_SIZE_SCALE) >= df.index[1]
    assert (trades['price'].max() -
            BIN_SIZE_SCALE) < df.index[-1] < trades['price'].max()

    assert 1667.0 == df.index[-1]

    # bid assert 763.7 == df['ask'].iat[0]  # ask
    assert 710.98 == df['bid_amount'].iat[0]
    assert 111 == df['bid'].iat[0]
    assert 52.71999999999997 == df['delta'].iat[0]  # delta
    # assert 50.659 == df['bid'].iat[-1]  # bid
    # assert 108.21 == df['ask'].iat[-1]  # ask
    # assert 57.551 == df['delta'].iat[-1]  # delta
    #

# bidask


def do_plot(pair, data, trades, plot_config=None):
    import plotly.offline as pyo

    from freqtrade.plot.plotting import generate_candlestick_graph

    # Filter trades to one pair
    trades_red = trades  # .loc[trades['pair'] == pair].copy()
    # Limit graph period to your BT timerange
    data_red = data  # data['2021-04-01':'2021-04-20']

    # plotconf = strategy.plot_config
    plotconf = plot_config

    # Generate candlestick graph
    graph = generate_candlestick_graph(pair=pair,
                                       data=data_red,
                                       trades=trades_red,
                                       plot_config=plotconf
                                       )
    pyo.plot(graph, output_type="file", show_link=False,
             filename="tests/data/test_converter_public_trades.html")


# need to be at last to see if some test changed the testdata
# always need to use .copy() to avoid changing the testdata
def test_public_trades_testdata_sanity(candles, trades, public_trades_list, public_trades_list_simple,
                                       public_trades_list_simple_bidask, public_trades_list_simple_results,
                                       populate_dataframe_with_trades_dataframe, populate_dataframe_with_trades_trades):
    assert 10999 == len(candles)
    assert 1811 == len(trades)
    assert 1000 == len(public_trades_list)
    assert 3 == len(public_trades_list_simple_results)
    assert 7 == len(public_trades_list_simple_bidask)
    assert 999 == len(populate_dataframe_with_trades_dataframe)
    assert 8033249 == len(populate_dataframe_with_trades_trades)

    assert 7 == len(public_trades_list_simple)
    assert 5 == public_trades_list_simple.loc[
        public_trades_list_simple['side'].str.contains(
            'sell'), 'id'].count()
    assert 2 == public_trades_list_simple.loc[
        public_trades_list_simple['side'].str.contains(
            'buy'), 'id'].count()

    assert public_trades_list.columns.tolist() == [
        'timestamp', 'id', 'type', 'side', 'price',
        'amount', 'cost', 'date']

    assert public_trades_list.columns.tolist() == [
        'timestamp', 'id', 'type', 'side', 'price', 'amount', 'cost', 'date']
    assert public_trades_list_simple_results.columns.tolist() == [
        'level', 'bid', 'ask', 'delta']
    assert public_trades_list_simple.columns.tolist() == [
        'timestamp', 'id', 'type', 'side', 'price',
        'amount', 'cost', 'date']
    assert populate_dataframe_with_trades_dataframe.columns.tolist() == [
        'date', 'open', 'high', 'low', 'close', 'volume']
    assert populate_dataframe_with_trades_trades.columns.tolist() == [
        'timestamp', 'id', 'type', 'side', 'price',
        'amount', 'cost', 'date']

    public_trades_list_simple_results = pd.DataFrame([[0, 0, 0, 0], [23437.5, 0.245, 0.0, -0.245], [23438.0, 0.0, 0.14, 0.140]],
                                                     columns=public_trades_list_simple_results.columns)
    pd.testing.assert_series_equal(
        public_trades_list_simple_results['delta'], public_trades_list_simple_results['delta'], check_index=False)
    assert public_trades_list_simple_results.values.tolist(
    ) == public_trades_list_simple_results.values.tolist()


class ReporterPlugin:
    def pytest_sessionfinish(self):
        print("*** test run reporting finishing")


# # invoke self to be able to debug
if __name__ == "__main__":
    import os
    import sys

    # print cwd
    print("cwd: ", os.getcwd())
    try:
        import pytest
        retval = pytest.main(
            ["--stepwise", "-k", "test_public_trades", "-vvv"], plugins=[ReporterPlugin()])
        sys.exit(retval)
    except ImportError:
        print("Please install pytest to run tests")
        sys.exit(1)
    except Exception as e:
        print(e)
        sys.exit(1)