def test_run_simulation_round_sell(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)[['close']]
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
ts = TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=None,
short_exposure_factor=1.5)
ts.runSimulationRound(i_index=0, signal=TRADE_SIGNALS['sell'])
ts.runSimulationRound(i_index=1, signal=TRADE_SIGNALS['sell'])
self.assertListEqual(list(ts._portfolio[1, :]),
[1.0, 1.0, 1.5 * ts._close_values[1, 0]])
self.assertListEqual(ts._simulation_data.iloc[1, :].to_list(), [
'sell', 'short', ts._close_values[1, 0],
ts._simulation_data['exposure'].iat[0] +
1.5 * ts._close_values[1, 0], -ts._close_values[1, 0], 0.0,
-ts._close_values[1, 0]
])
def test_close_simulation(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)[['close']]
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
ts = TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=None,
short_exposure_factor=1.5)
self.maxDiff = None
statistics_expected_result = {
'number_of_trading_days': 3169,
'number_of_buy_signals': 1297,
'number_of_ignored_buy_signals': 1029,
'number_of_sell_signals': 1152,
'number_of_ignored_sell_signals': 741,
'last_stock_value': 140.41,
'last_exposure': 3268.0,
'last_open_long_positions': 397,
'last_open_short_positions': 396,
'last_portfolio_value': 3368.0,
'last_earnings': 3468.0,
'final_balance': 3568.0
}
ts._simulation_data = pd.read_csv(
'./data/simulation_data_full_with_actions.csv',
parse_dates=True,
index_col=0)
ts._portfolio = pd.read_csv(
'./data/portfolio_simulation_data_full.csv',
parse_dates=True,
index_col=0).to_numpy(dtype=np.float64, copy=True)
sd_result, st_result = ts.closeSimulation()
pd.testing.assert_frame_equal(sd_result, ts._simulation_data)
self.assertDictEqual(st_result, statistics_expected_result)
def test_close_values_missing(self):
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
with self.assertRaises(TypeError):
TradingSimulation(input_data_index=id_df.index,
max_exposure=None,
short_exposure_factor=1.5)
def test_input_data_index_missing(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)[['close']]
with self.assertRaises(TypeError):
TradingSimulation(close_values=cv_df,
max_exposure=None,
short_exposure_factor=1.5)
def test_run_simulation_round_buy_max_exposure(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)[['close']]
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
ts = TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=1,
short_exposure_factor=1.5)
ts.runSimulationRound(i_index=0, signal=TRADE_SIGNALS['buy'])
ts.runSimulationRound(i_index=1, signal=TRADE_SIGNALS['buy'])
ts.runSimulationRound(i_index=2, signal=TRADE_SIGNALS['buy'])
self.assertListEqual(list(ts._portfolio[2, :]), [0.0, 0.0, 0.0])
self.assertListEqual(
ts._simulation_data.iloc[2, :].to_list(),
['buy', 'none', ts._close_values[2, 0], 0.0, 0.0, 0.0, 0.0])
def test_close_open_positions_none_open_no_force_all_write(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)[['close']]
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
ts = TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=None,
short_exposure_factor=1.5)
ts._portfolio = pd.read_csv(
'./data/portfolio_simulation_data_none_open_ten_rounds.csv',
parse_dates=True,
index_col=0).to_numpy(dtype=np.float64, copy=True)
portfolio_expected_result = pd.read_csv(
'./data/portfolio_simulation_data_none_open_ten_rounds.csv',
parse_dates=True,
index_col=0).to_numpy(dtype=np.float64, copy=True)
ts._simulation_data['exposure'].iat[8] = 100.0
ts._simulation_data['earnings'].iat[8] = 200.0
earnings, closed_exposure = ts._closeOpenPositions(i_index=9)
self.assertEqual(earnings, 0.0)
self.assertEqual(closed_exposure, 0.0)
np.testing.assert_equal(ts._simulation_data['exposure'].iat[9], 100.0)
np.testing.assert_equal(ts._simulation_data['earnings'].iat[9], 200.0)
np.testing.assert_equal(ts._portfolio, portfolio_expected_result)
def test_close_open_positions_with_open_no_force_all_write(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)[['close']]
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
ts = TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=None,
short_exposure_factor=1.5)
ts._portfolio = pd.read_csv(
'./data/portfolio_simulation_data_all_open_ten_rounds.csv',
parse_dates=True,
index_col=0).to_numpy(dtype=np.float64, copy=True)
portfolio_expected_result = pd.read_csv(
'./data/portfolio_simulation_data_all_open_ten_rounds.csv',
parse_dates=True,
index_col=0).to_numpy(dtype=np.float64, copy=True)
portfolio_expected_result[0, 1] = 2.0
portfolio_expected_result[2, 1] = 2.0
portfolio_expected_result[5, 1] = 2.0
portfolio_expected_result[5, 2] = 40.00
ts._simulation_data['exposure'].iat[8] = 100.0
ts._simulation_data['earnings'].iat[8] = 200.0
ts._portfolio[5, 2] = 40.00
ts._close_values[9, 0] = 24.0
earnings, closed_exposure = ts._closeOpenPositions(i_index=9)
exposure_expected = 20.50 + 22.50 + 40.00
earnings_expected = (2 * 24.0 - 20.5 - 22.5) + \
((40.00 / 1.5) - 1 * 24.0)
self.assertEqual(earnings, earnings_expected)
self.assertEqual(closed_exposure, exposure_expected)
np.testing.assert_equal(
ts._simulation_data['exposure'].iat[9],
ts._simulation_data['exposure'].iat[8] - exposure_expected)
np.testing.assert_equal(
ts._simulation_data['earnings'].iat[9],
ts._simulation_data['earnings'].iat[8] + earnings_expected)
np.testing.assert_equal(ts._portfolio, portfolio_expected_result)
def test_close_values_wrong_type(self):
# close values DataFrame
cv_df = 'No-DF'
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
with self.assertRaises(NotValidInputDataForSimulation):
TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=None,
short_exposure_factor=1.5)
def test_calculate_portfolio_value(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)[['close']]
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
ts = TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=None,
short_exposure_factor=1.5)
ts._portfolio = pd.read_csv(
'./data/portfolio_simulation_data_ten_rounds.csv',
parse_dates=True,
index_col=0).to_numpy(dtype=np.float64, copy=True)
value = ts._calculatePortfolioValue(i_index=9)
self.assertEqual(value, 34.37 * (2 - 3))
def test_short_exposure_factor_below_one(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)[['close']]
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
with self.assertRaises(WrongValueForInputParameter):
TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=None,
short_exposure_factor=0.99)
def test_close_values_missing_column(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0).drop(columns=['close'], axis=1)
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
with self.assertRaises(NotValidInputDataForSimulation):
TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=None,
short_exposure_factor=1.5)
def test_short_exposure_factor_missing(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)[['close']]
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
ts = TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=None)
self.assertEqual(ts._short_exposure_factor, 1.5)
def test_calculate_simulation_statistics_ten_simulation_rounds(self):
# close values DataFrame
cv_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)[['close']]
# input_data DataFrame
id_df = pd.read_csv('./data/sample_data.csv',
parse_dates=True,
index_col=0)
ts = TradingSimulation(input_data_index=id_df.index,
close_values=cv_df,
max_exposure=None,
short_exposure_factor=1.5)
statistics_expected_result = {
'number_of_trading_days': 10,
'number_of_buy_signals': 4,
'number_of_ignored_buy_signals': 1,
'number_of_sell_signals': 3,
'number_of_ignored_sell_signals': 2,
'last_stock_value': 34.37,
'last_exposure': 109.0,
'last_open_long_positions': 2,
'last_open_short_positions': 3,
'last_portfolio_value': 209.0,
'last_earnings': 309.0,
'final_balance': 409.0
}
ts._simulation_data = pd.read_csv(
'./data/simulation_data_with_actions_ten_rounds.csv',
parse_dates=True,
index_col=0)
ts._portfolio = pd.read_csv(
'./data/portfolio_simulation_data_ten_rounds.csv',
parse_dates=True,
index_col=0).to_numpy(dtype=np.float64, copy=True)
ts._calculateSimulationStatistics()
self.assertDictEqual(ts._statistics, statistics_expected_result)