def test_marketsim(description, group, inputs, outputs, grader):
"""Test compute_portvals() returns correct daily portfolio values.
Requires test description, test case group, inputs, expected outputs, and a grader fixture.
"""
points_earned = 0.0 # initialize points for this test case
try:
# Try to import student code (only once)
if not main_code in globals():
import importlib
# * Import module
mod = importlib.import_module(main_code)
globals()[main_code] = mod
# * Import methods to test
for m in ['compute_portvals']:
globals()[m] = getattr(mod, m)
incorrect = False
msgs = []
if group == 'author':
try:
auth_string = run_with_timeout(marketsim.author,
seconds_per_test_case, (), {})
if auth_string == 'tb34':
incorrect = True
msgs.append(" Incorrect author name (tb34)")
points_earned = -10
elif auth_string == '':
incorrect = True
msgs.append(" Empty author name")
points_earned = -10
except Exception as e:
incorrect = True
msgs.append(
" Exception occured when calling author() method: {}".
format(e))
points_earned = -10
else:
# Unpack test case
orders_file = inputs['orders_file']
start_val = inputs['start_val']
impct = inputs['impact']
commish = inputs['commission']
portvals = None
fullpath_orders_file = get_orders_data_file(orders_file)
portvals = run_with_timeout(
compute_portvals, seconds_per_test_case, (), {
'orders_file': fullpath_orders_file,
'start_val': start_val,
'commission': commish,
'impact': impct
})
# * Check return type is correct, coax into Series
assert (type(portvals) == pd.Series) or (
type(portvals) == pd.DataFrame and len(portvals.columns)
== 1), "You must return a Series or single-column DataFrame!"
if type(portvals) == pd.DataFrame:
portvals = portvals[portvals.columns[
0]] # convert single-column DataFrame to Series
if group == 'basic':
if len(portvals) != outputs['num_days']:
incorrect = True
msgs.append(
" Incorrect number of days: {}, expected {}".format(
len(portvals), outputs['num_days']))
else:
points_earned += 2.0
if abs(portvals[-1] - outputs['last_day_portval']) > (
0.001 * outputs['last_day_portval']):
incorrect = True
msgs.append(
" Incorrect final value: {}, expected {}".format(
portvals[-1], outputs['last_day_portval']))
else:
points_earned += 5.0
adr, sr = get_stats(portvals)
if abs(sr - outputs['sharpe_ratio']) > abs(
0.001 * outputs['sharpe_ratio']):
incorrect = True
msgs.append(
" Incorrect sharpe ratio: {}, expected {}".format(
sr, outputs['sharpe_ratio']))
else:
points_earned += 1.0
if abs(adr - outputs['avg_daily_ret']) > abs(
0.001 * outputs['avg_daily_ret']):
incorrect = True
msgs.append(
" Incorrect avg daily return: {}, expected {}".
format(adr, outputs['avg_daily_ret']))
else:
points_earned += 1.0
elif group == 'commission' or group == 'impact' or group == 'both':
if abs(portvals[-1] - outputs['last_day_portval']) > 0.001:
incorrect = True
msgs.append(
" Incorrect final value: {}, expected {}".format(
portvals[-1], outputs['last_day_portval']))
else:
points_earned += 2.0
if incorrect:
raise IncorrectOutput, "Test failed on one or more output criteria.\n Inputs:\n{}\n Failures:\n{}".format(
inputs, "\n".join(msgs))
except Exception as e:
# Test result: failed
msg = "Test case description: {}\n".format(description)
# Generate a filtered stacktrace, only showing erroneous lines in student file(s)
tb_list = tb.extract_tb(sys.exc_info()[2])
if 'grading_traceback' in dir(e):
tb_list = e.grading_traceback
for i in xrange(len(tb_list)):
row = tb_list[i]
tb_list[i] = (os.path.basename(row[0]), row[1], row[2], row[3]
) # show only filename instead of long absolute path
tb_list = [row for row in tb_list if row[0] == 'marketsim_old.py']
if tb_list:
msg += "Traceback:\n"
msg += ''.join(tb.format_list(tb_list)) # contains newlines
msg += "{}: {}".format(e.__class__.__name__, e.message)
# Report failure result to grader, with stacktrace
grader.add_result(
GradeResult(outcome='failed',
points=max(points_earned, 0),
msg=msg))
raise
else:
# Test result: passed (no exceptions)
grader.add_result(
GradeResult(outcome='passed', points=points_earned, msg=None))
def compute_portvals(orders_file="./orders/orders.csv", start_val=1000000):
orders = get_orders_data_file(orders_file)
orders_data = pd.read_csv(orders)
orders_data.sort_values(by='Date', inplace=True, ascending=True)
start_date = orders_data.iloc[0, 0]
end_date = orders_data.iloc[len(orders_data) - 1, 0]
adj_close_prices = get_data(['SPY'], pd.date_range(start_date, end_date))
orders_data_new = orders_data.set_index('Date', inplace=False)
orders_join = orders_data_new.join(adj_close_prices)
adj_close_prices = adj_close_prices.join(orders_data_new)
adj_close_prices = adj_close_prices[['SPY']]
adj_close_prices.rename(columns={'SPY': 'Portval'}, inplace=True)
portvals = pd.DataFrame().reindex_like(adj_close_prices)
cash = start_val
equity = 0
portval = cash + equity
shares_allocation = {}
daily_cash = pd.DataFrame().reindex_like(adj_close_prices)
daily_shares_allocation = {}
df = adj_close_prices.index
df1 = orders_join.index
i = 0
orders_data_index = 0
value = df[0]
while value in df:
value1 = pd.to_datetime(value).strftime('%Y-%m-%d')
if (value in df1):
sd = orders_data.iloc[orders_data_index, 0]
sym = orders_data.iloc[orders_data_index, 1]
order = orders_data.iloc[orders_data_index, 2]
shares = orders_data.iloc[orders_data_index, 3]
adj_close_price = get_data([sym], pd.date_range(sd, sd))
adj_close_price = adj_close_price[[sym]]
commission = 9.95
if (not np.isnan(adj_close_price.iloc[0, 0])):
if (order == 'SELL'):
if (i != 0):
cash = daily_cash.iloc[i - 1, 0]
cash += adj_close_price.iloc[0, 0] * 0.995 * shares
cash -= commission
if (i != 0):
previous_date = pd.to_datetime(
df[i - 1]).strftime('%Y-%m-%d')
shares_allocation = daily_shares_allocation[
previous_date]
shares_allocation_n = shares_allocation.copy()
if (shares_allocation_n.has_key(sym)):
shares_allocation_n[sym] -= shares
else:
shares_allocation_n[sym] = 0 - shares
if (order == 'BUY'):
if (i != 0):
cash = daily_cash.iloc[i - 1, 0]
cash -= adj_close_price.iloc[0, 0] * 1.005 * shares
cash -= commission
if (i != 0):
previous_date = pd.to_datetime(
df[i - 1]).strftime('%Y-%m-%d')
shares_allocation = daily_shares_allocation[
previous_date]
shares_allocation_n = shares_allocation.copy()
if (not shares_allocation.has_key(sym)):
shares_allocation_n[sym] = shares
else:
shares_allocation_n[sym] += shares
orders_data_index += 1
leverage = sum_of_abs_of_stock_positions(
shares_allocation_n,
value1) / (compute_equity(shares_allocation_n, value1) + cash)
if (leverage <= 2.0):
daily_shares_allocation[value1] = shares_allocation_n
daily_cash.set_value(value1, 'Portval', cash)
equity = compute_equity(shares_allocation_n, value1)
portval = cash + equity
portvals.set_value(sd, 'Portval', portval)
else:
last_date = pd.to_datetime(df[i - 1]).strftime('%Y-%m-%d')
sa = daily_shares_allocation[last_date]
daily_shares_allocation[value1] = sa
c = daily_cash.iloc[i - 1, 0]
daily_cash.set_value(value1, 'Portval', c)
e = compute_equity(sa, value1)
p = c + e
portvals.set_value(sd, 'Portval', p)
i += 1
if (i == len(df)):
break
value = df[i]
else:
current_date = value1
previous_date = pd.to_datetime(df[i - 1]).strftime('%Y-%m-%d')
next_order_date = pd.to_datetime(
df1[orders_data_index]).strftime('%Y-%m-%d')
new_data = get_data(['SPY'],
pd.date_range(current_date, next_order_date))
l = i + len(new_data) - 2
b = i
date_before_next_order_date = pd.to_datetime(
df[l]).strftime('%Y-%m-%d')
cash_new = daily_cash.iloc[i - 1, 0]
shares_allocation_new = daily_shares_allocation[previous_date]
equity = compute_equity_range(shares_allocation_new, current_date,
date_before_next_order_date)
portvals_intermed = equity + cash_new
portvals['Portval'][
current_date:date_before_next_order_date] = portvals_intermed
daily_cash['Portval'][
current_date:date_before_next_order_date] = cash_new
for a in range(b, l + 1):
date_assigned = pd.to_datetime(df[a]).strftime('%Y-%m-%d')
daily_shares_allocation[date_assigned] = shares_allocation_new
i = l + 1
value = df[i]
portvals = portvals.groupby(portvals.index).last()
return portvals