freqtrade_origin/tests/exchange/test_exchange_utils.py

344 lines
14 KiB
Python

# pragma pylint: disable=missing-docstring, protected-access, invalid-name
from datetime import datetime, timedelta, timezone
import pytest
from ccxt import (DECIMAL_PLACES, ROUND, ROUND_DOWN, ROUND_UP, SIGNIFICANT_DIGITS, TICK_SIZE,
TRUNCATE)
from freqtrade.enums import RunMode
from freqtrade.exceptions import OperationalException
from freqtrade.exchange import (amount_to_contract_precision, amount_to_precision,
date_minus_candles, price_to_precision, timeframe_to_minutes,
timeframe_to_msecs, timeframe_to_next_date, timeframe_to_prev_date,
timeframe_to_resample_freq, timeframe_to_seconds)
from freqtrade.exchange.check_exchange import check_exchange
from tests.conftest import log_has_re
def test_check_exchange(default_conf, caplog) -> None:
# Test an officially supported by Freqtrade team exchange
default_conf['runmode'] = RunMode.DRY_RUN
default_conf.get('exchange').update({'name': 'BINANCE'})
assert check_exchange(default_conf)
assert log_has_re(r"Exchange .* is officially supported by the Freqtrade development team\.",
caplog)
caplog.clear()
# Test an officially supported by Freqtrade team exchange
default_conf.get('exchange').update({'name': 'binance'})
assert check_exchange(default_conf)
assert log_has_re(
r"Exchange \"binance\" is officially supported by the Freqtrade development team\.",
caplog)
caplog.clear()
# Test an officially supported by Freqtrade team exchange
default_conf.get('exchange').update({'name': 'binanceus'})
assert check_exchange(default_conf)
assert log_has_re(
r"Exchange \"binanceus\" is officially supported by the Freqtrade development team\.",
caplog)
caplog.clear()
# Test an officially supported by Freqtrade team exchange - with remapping
default_conf.get('exchange').update({'name': 'okx'})
assert check_exchange(default_conf)
assert log_has_re(
r"Exchange \"okx\" is officially supported by the Freqtrade development team\.",
caplog)
caplog.clear()
# Test an available exchange, supported by ccxt
default_conf.get('exchange').update({'name': 'huobijp'})
assert check_exchange(default_conf)
assert log_has_re(r"Exchange .* is known to the the ccxt library, available for the bot, "
r"but not officially supported "
r"by the Freqtrade development team\. .*", caplog)
caplog.clear()
# Test a 'bad' exchange, which known to have serious problems
default_conf.get('exchange').update({'name': 'bitmex'})
with pytest.raises(OperationalException,
match=r"Exchange .* will not work with Freqtrade\..*"):
check_exchange(default_conf)
caplog.clear()
# Test a 'bad' exchange with check_for_bad=False
default_conf.get('exchange').update({'name': 'bitmex'})
assert check_exchange(default_conf, False)
assert log_has_re(r"Exchange .* is known to the the ccxt library, available for the bot, "
r"but not officially supported "
r"by the Freqtrade development team\. .*", caplog)
caplog.clear()
# Test an invalid exchange
default_conf.get('exchange').update({'name': 'unknown_exchange'})
with pytest.raises(
OperationalException,
match=r'Exchange "unknown_exchange" is not known to the ccxt library '
r'and therefore not available for the bot.*'
):
check_exchange(default_conf)
# Test no exchange...
default_conf.get('exchange').update({'name': ''})
default_conf['runmode'] = RunMode.PLOT
assert check_exchange(default_conf)
# Test no exchange...
default_conf.get('exchange').update({'name': ''})
default_conf['runmode'] = RunMode.UTIL_EXCHANGE
with pytest.raises(OperationalException,
match=r'This command requires a configured exchange.*'):
check_exchange(default_conf)
def test_date_minus_candles():
date = datetime(2019, 8, 12, 13, 25, 0, tzinfo=timezone.utc)
assert date_minus_candles("5m", 3, date) == date - timedelta(minutes=15)
assert date_minus_candles("5m", 5, date) == date - timedelta(minutes=25)
assert date_minus_candles("1m", 6, date) == date - timedelta(minutes=6)
assert date_minus_candles("1h", 3, date) == date - timedelta(hours=3, minutes=25)
assert date_minus_candles("1h", 3) == timeframe_to_prev_date('1h') - timedelta(hours=3)
def test_timeframe_to_minutes():
assert timeframe_to_minutes("5m") == 5
assert timeframe_to_minutes("10m") == 10
assert timeframe_to_minutes("1h") == 60
assert timeframe_to_minutes("1d") == 1440
def test_timeframe_to_seconds():
assert timeframe_to_seconds("5m") == 300
assert timeframe_to_seconds("10m") == 600
assert timeframe_to_seconds("1h") == 3600
assert timeframe_to_seconds("1d") == 86400
def test_timeframe_to_msecs():
assert timeframe_to_msecs("5m") == 300000
assert timeframe_to_msecs("10m") == 600000
assert timeframe_to_msecs("1h") == 3600000
assert timeframe_to_msecs("1d") == 86400000
@pytest.mark.parametrize("timeframe,expected", [
("1s", '1s'),
("15s", '15s'),
("5m", '300s'),
("10m", '600s'),
("1h", '3600s'),
("1d", '86400s'),
("1w", '1W-MON'),
("1M", '1MS'),
("1y", '1YS'),
])
def test_timeframe_to_resample_freq(timeframe, expected):
assert timeframe_to_resample_freq(timeframe) == expected
def test_timeframe_to_prev_date():
# 2019-08-12 13:22:08
date = datetime.fromtimestamp(1565616128, tz=timezone.utc)
tf_list = [
# 5m -> 2019-08-12 13:20:00
("5m", datetime(2019, 8, 12, 13, 20, 0, tzinfo=timezone.utc)),
# 10m -> 2019-08-12 13:20:00
("10m", datetime(2019, 8, 12, 13, 20, 0, tzinfo=timezone.utc)),
# 1h -> 2019-08-12 13:00:00
("1h", datetime(2019, 8, 12, 13, 00, 0, tzinfo=timezone.utc)),
# 2h -> 2019-08-12 12:00:00
("2h", datetime(2019, 8, 12, 12, 00, 0, tzinfo=timezone.utc)),
# 4h -> 2019-08-12 12:00:00
("4h", datetime(2019, 8, 12, 12, 00, 0, tzinfo=timezone.utc)),
# 1d -> 2019-08-12 00:00:00
("1d", datetime(2019, 8, 12, 00, 00, 0, tzinfo=timezone.utc)),
]
for interval, result in tf_list:
assert timeframe_to_prev_date(interval, date) == result
date = datetime.now(tz=timezone.utc)
assert timeframe_to_prev_date("5m") < date
# Does not round
time = datetime(2019, 8, 12, 13, 20, 0, tzinfo=timezone.utc)
assert timeframe_to_prev_date('5m', time) == time
time = datetime(2019, 8, 12, 13, 0, 0, tzinfo=timezone.utc)
assert timeframe_to_prev_date('1h', time) == time
def test_timeframe_to_next_date():
# 2019-08-12 13:22:08
date = datetime.fromtimestamp(1565616128, tz=timezone.utc)
tf_list = [
# 5m -> 2019-08-12 13:25:00
("5m", datetime(2019, 8, 12, 13, 25, 0, tzinfo=timezone.utc)),
# 10m -> 2019-08-12 13:30:00
("10m", datetime(2019, 8, 12, 13, 30, 0, tzinfo=timezone.utc)),
# 1h -> 2019-08-12 14:00:00
("1h", datetime(2019, 8, 12, 14, 00, 0, tzinfo=timezone.utc)),
# 2h -> 2019-08-12 14:00:00
("2h", datetime(2019, 8, 12, 14, 00, 0, tzinfo=timezone.utc)),
# 4h -> 2019-08-12 14:00:00
("4h", datetime(2019, 8, 12, 16, 00, 0, tzinfo=timezone.utc)),
# 1d -> 2019-08-13 00:00:00
("1d", datetime(2019, 8, 13, 0, 0, 0, tzinfo=timezone.utc)),
]
for interval, result in tf_list:
assert timeframe_to_next_date(interval, date) == result
date = datetime.now(tz=timezone.utc)
assert timeframe_to_next_date("5m") > date
date = datetime(2019, 8, 12, 13, 30, 0, tzinfo=timezone.utc)
assert timeframe_to_next_date("5m", date) == date + timedelta(minutes=5)
@pytest.mark.parametrize("amount,precision_mode,precision,expected", [
(2.34559, DECIMAL_PLACES, 4, 2.3455),
(2.34559, DECIMAL_PLACES, 5, 2.34559),
(2.34559, DECIMAL_PLACES, 3, 2.345),
(2.9999, DECIMAL_PLACES, 3, 2.999),
(2.9909, DECIMAL_PLACES, 3, 2.990),
(2.9909, DECIMAL_PLACES, 0, 2),
(29991.5555, DECIMAL_PLACES, 0, 29991),
(29991.5555, DECIMAL_PLACES, -1, 29990),
(29991.5555, DECIMAL_PLACES, -2, 29900),
# Tests for
(2.34559, SIGNIFICANT_DIGITS, 4, 2.345),
(2.34559, SIGNIFICANT_DIGITS, 5, 2.3455),
(2.34559, SIGNIFICANT_DIGITS, 3, 2.34),
(2.9999, SIGNIFICANT_DIGITS, 3, 2.99),
(2.9909, SIGNIFICANT_DIGITS, 3, 2.99),
(0.0000077723, SIGNIFICANT_DIGITS, 5, 0.0000077723),
(0.0000077723, SIGNIFICANT_DIGITS, 3, 0.00000777),
(0.0000077723, SIGNIFICANT_DIGITS, 1, 0.000007),
# Tests for Tick-size
(2.34559, TICK_SIZE, 0.0001, 2.3455),
(2.34559, TICK_SIZE, 0.00001, 2.34559),
(2.34559, TICK_SIZE, 0.001, 2.345),
(2.9999, TICK_SIZE, 0.001, 2.999),
(2.9909, TICK_SIZE, 0.001, 2.990),
(2.9909, TICK_SIZE, 0.005, 2.99),
(2.9999, TICK_SIZE, 0.005, 2.995),
])
def test_amount_to_precision(amount, precision_mode, precision, expected,):
"""
Test rounds down
"""
# digits counting mode
# DECIMAL_PLACES = 2
# SIGNIFICANT_DIGITS = 3
# TICK_SIZE = 4
assert amount_to_precision(amount, precision, precision_mode) == expected
@pytest.mark.parametrize("price,precision_mode,precision,expected,rounding_mode", [
# Tests for DECIMAL_PLACES, ROUND_UP
(2.34559, DECIMAL_PLACES, 4, 2.3456, ROUND_UP),
(2.34559, DECIMAL_PLACES, 5, 2.34559, ROUND_UP),
(2.34559, DECIMAL_PLACES, 3, 2.346, ROUND_UP),
(2.9999, DECIMAL_PLACES, 3, 3.000, ROUND_UP),
(2.9909, DECIMAL_PLACES, 3, 2.991, ROUND_UP),
(2.9901, DECIMAL_PLACES, 3, 2.991, ROUND_UP),
(2.34559, DECIMAL_PLACES, 5, 2.34559, ROUND_DOWN),
(2.34559, DECIMAL_PLACES, 4, 2.3455, ROUND_DOWN),
(2.9901, DECIMAL_PLACES, 3, 2.990, ROUND_DOWN),
(0.00299, DECIMAL_PLACES, 3, 0.002, ROUND_DOWN),
# Tests for DECIMAL_PLACES, ROUND
(2.345600000000001, DECIMAL_PLACES, 4, 2.3456, ROUND),
(2.345551, DECIMAL_PLACES, 4, 2.3456, ROUND),
(2.49, DECIMAL_PLACES, 0, 2., ROUND),
(2.51, DECIMAL_PLACES, 0, 3., ROUND),
(5.1, DECIMAL_PLACES, -1, 10., ROUND),
(4.9, DECIMAL_PLACES, -1, 0., ROUND),
(0.000007222, SIGNIFICANT_DIGITS, 1, 0.000007, ROUND),
(0.000007222, SIGNIFICANT_DIGITS, 2, 0.0000072, ROUND),
(0.000007777, SIGNIFICANT_DIGITS, 2, 0.0000078, ROUND),
# Tests for TICK_SIZE, ROUND_UP
(2.34559, TICK_SIZE, 0.0001, 2.3456, ROUND_UP),
(2.34559, TICK_SIZE, 0.00001, 2.34559, ROUND_UP),
(2.34559, TICK_SIZE, 0.001, 2.346, ROUND_UP),
(2.9999, TICK_SIZE, 0.001, 3.000, ROUND_UP),
(2.9909, TICK_SIZE, 0.001, 2.991, ROUND_UP),
(2.9909, TICK_SIZE, 0.001, 2.990, ROUND_DOWN),
(2.9909, TICK_SIZE, 0.005, 2.995, ROUND_UP),
(2.9973, TICK_SIZE, 0.005, 3.0, ROUND_UP),
(2.9977, TICK_SIZE, 0.005, 3.0, ROUND_UP),
(234.43, TICK_SIZE, 0.5, 234.5, ROUND_UP),
(234.43, TICK_SIZE, 0.5, 234.0, ROUND_DOWN),
(234.53, TICK_SIZE, 0.5, 235.0, ROUND_UP),
(234.53, TICK_SIZE, 0.5, 234.5, ROUND_DOWN),
(0.891534, TICK_SIZE, 0.0001, 0.8916, ROUND_UP),
(64968.89, TICK_SIZE, 0.01, 64968.89, ROUND_UP),
(0.000000003483, TICK_SIZE, 1e-12, 0.000000003483, ROUND_UP),
# Tests for TICK_SIZE, ROUND
(2.49, TICK_SIZE, 1., 2., ROUND),
(2.51, TICK_SIZE, 1., 3., ROUND),
(2.000000051, TICK_SIZE, 0.0000001, 2.0000001, ROUND),
(2.000000049, TICK_SIZE, 0.0000001, 2., ROUND),
(2.9909, TICK_SIZE, 0.005, 2.990, ROUND),
(2.9973, TICK_SIZE, 0.005, 2.995, ROUND),
(2.9977, TICK_SIZE, 0.005, 3.0, ROUND),
(234.24, TICK_SIZE, 0.5, 234., ROUND),
(234.26, TICK_SIZE, 0.5, 234.5, ROUND),
# Tests for TRUNCATTE
(2.34559, DECIMAL_PLACES, 4, 2.3455, TRUNCATE),
(2.34559, DECIMAL_PLACES, 5, 2.34559, TRUNCATE),
(2.34559, DECIMAL_PLACES, 3, 2.345, TRUNCATE),
(2.9999, DECIMAL_PLACES, 3, 2.999, TRUNCATE),
(2.9909, DECIMAL_PLACES, 3, 2.990, TRUNCATE),
(2.9909, TICK_SIZE, 0.001, 2.990, TRUNCATE),
(2.9909, TICK_SIZE, 0.01, 2.99, TRUNCATE),
(2.9909, TICK_SIZE, 0.1, 2.9, TRUNCATE),
# Tests for Significant
(2.34559, SIGNIFICANT_DIGITS, 4, 2.345, TRUNCATE),
(2.34559, SIGNIFICANT_DIGITS, 5, 2.3455, TRUNCATE),
(2.34559, SIGNIFICANT_DIGITS, 3, 2.34, TRUNCATE),
(2.9999, SIGNIFICANT_DIGITS, 3, 2.99, TRUNCATE),
(2.9909, SIGNIFICANT_DIGITS, 2, 2.9, TRUNCATE),
(0.00000777, SIGNIFICANT_DIGITS, 2, 0.0000077, TRUNCATE),
(0.00000729, SIGNIFICANT_DIGITS, 2, 0.0000072, TRUNCATE),
# ROUND
(722.2, SIGNIFICANT_DIGITS, 1, 700.0, ROUND),
(790.2, SIGNIFICANT_DIGITS, 1, 800.0, ROUND),
(722.2, SIGNIFICANT_DIGITS, 2, 720.0, ROUND),
(722.2, SIGNIFICANT_DIGITS, 1, 800.0, ROUND_UP),
(722.2, SIGNIFICANT_DIGITS, 2, 730.0, ROUND_UP),
(777.7, SIGNIFICANT_DIGITS, 2, 780.0, ROUND_UP),
(777.7, SIGNIFICANT_DIGITS, 3, 778.0, ROUND_UP),
(722.2, SIGNIFICANT_DIGITS, 1, 700.0, ROUND_DOWN),
(722.2, SIGNIFICANT_DIGITS, 2, 720.0, ROUND_DOWN),
(777.7, SIGNIFICANT_DIGITS, 2, 770.0, ROUND_DOWN),
(777.7, SIGNIFICANT_DIGITS, 3, 777.0, ROUND_DOWN),
(0.000007222, SIGNIFICANT_DIGITS, 1, 0.000008, ROUND_UP),
(0.000007222, SIGNIFICANT_DIGITS, 2, 0.0000073, ROUND_UP),
(0.000007777, SIGNIFICANT_DIGITS, 2, 0.0000078, ROUND_UP),
(0.000007222, SIGNIFICANT_DIGITS, 1, 0.000007, ROUND_DOWN),
(0.000007222, SIGNIFICANT_DIGITS, 2, 0.0000072, ROUND_DOWN),
(0.000007777, SIGNIFICANT_DIGITS, 2, 0.0000077, ROUND_DOWN),
])
def test_price_to_precision(price, precision_mode, precision, expected, rounding_mode):
assert price_to_precision(
price, precision, precision_mode, rounding_mode=rounding_mode) == expected
@pytest.mark.parametrize('amount,precision,precision_mode,contract_size,expected', [
(1.17, 1.0, 4, 0.01, 1.17), # Tick size
(1.17, 1.0, 2, 0.01, 1.17), #
(1.16, 1.0, 4, 0.01, 1.16), #
(1.16, 1.0, 2, 0.01, 1.16), #
(1.13, 1.0, 2, 0.01, 1.13), #
(10.988, 1.0, 2, 10, 10),
(10.988, 1.0, 4, 10, 10),
])
def test_amount_to_contract_precision_standalone(amount, precision, precision_mode, contract_size,
expected):
res = amount_to_contract_precision(amount, precision, precision_mode, contract_size)
assert pytest.approx(res) == expected