def eval_scenario(returns_params, backtest_params, portfolio_params):
excess_returns = data.load_excess_returns(**returns_params)
# Setup backtest return splits.
training_returns, testing_returns = (
backtest.split_returns(excess_returns, **backtest_params))
results = list()
period = 0
for er_train, er_test in zip(training_returns, testing_returns):
sol = mvportfolio.portfolio(
er_train,
**portfolio_params)
portfolio = sol
returns = er_test.dot(portfolio)
result = {'portfolio': array2str(portfolio)}
result['test_returns'] = array2str(returns)
result['period'] = period
period += 1
results.append(result)
logging.debug('period = {0}'.format(period))
return {'kwargs': {'returns': returns_params,
'backtest': backtest_params,
'portfolio': portfolio_params},
'results': results}
def eval_scenario(returns_params, backtest_params, portfolio_params):
excess_returns = data.load_excess_returns(**returns_params)
# Setup backtest return splits.
training_returns, testing_returns = (
backtest.split_returns(excess_returns, **backtest_params))
results = list()
period = 0
for er_train, er_test in zip(training_returns, testing_returns):
try:
sol = dccportfolio.portfolio_sphere(
er_train,
extended_return=True, verbose=False,
**portfolio_params)
except LinAlgError:
logging.error('Linear algebra error encountered.')
sol = ('', '', '', '', 'linear_algebra_error')
portfolio = sol[0]
status = sol[4]
if status == 'solution_found':
returns = er_test.dot(portfolio)
else:
returns = ''
sol_keys = ['portfolio', 'objective', 'sdp_solution', 'constraints', 'status']
result = {
k: pg.array2str(s)
for k, s in zip(sol_keys, sol)
}
result['test_returns'] = pg.array2str(returns)
result['period'] = period
period += 1
results.append(result)
logging.debug('period = {0}'.format(period))
return {'kwargs': {'returns': returns_params,
'backtest': backtest_params,
'portfolio': portfolio_params},
'results': results}
backtest_results = pd.read_sql(query, db.conn)
scenarios = backtest_results.groupby(
['experiment_id', 'backtest_id', 'parameters_id'])
outputs = list()
for bt in scenarios:
print(bt[0])
experiment_id, backtest_id, parameters_id = bt[0]
results = bt[1]
experiment = db.load_experiment(experiment_id)
backtest_params = db.load_backtests([backtest_id])
returns = db.load_returns(experiment)
excess_returns = returns.values
train, test = backtest.split_returns(
excess_returns,
backtest_params['update_periods'],
backtest_params['trailing_periods'])
portfolios = pg.decimallistseries2nparray(results.portfolio_weights)
portfolio_returns = pg.decimallistseries2nparray(results.test_returns)
train_mean = ps.mean(portfolios[0,:], train[0])
train_variance = ps.variance(portfolios[0,:], train[0])
eta = results.eta.iloc[0]
backtest_statistics = ps.rkw_statistics(
portfolios, portfolio_returns, eta=eta)
test_statistics = backtest_statistics['statistics']
agg_returns = backtest_statistics['aggregate_returns']
output = test_statistics
output['aggregate_returns'] = pg.array2pgarray(agg_returns)
output['train_mean_return'] = train_mean
output['train_variance'] = train_variance
