def test_validate_date():
bad_dates = ["2020-02-", "", None, "20-02-02", "2020-02-1"]
for d in bad_dates:
with pytest.raises(AssertionError):
validate_date(d)
good_dates = ["2020-02-02", "2020-02-01", "2021-12-31"]
for d in good_dates:
validate_date(d)
def show_all_stocks(request):
all_dates = all_available_dates()
if len(all_dates) < 1:
raise Http404("No ASX price data available!")
ymd = all_dates[-1]
validate_date(ymd)
qs = valid_quotes_only(ymd)
timeframe = Timeframe()
return show_companies(qs, request, timeframe, extra_context={
"title": "All stocks",
"sentiment_heatmap_title": "All stock sentiment: {}".format(timeframe.description)
})
def test_timeframe(data, expected):
#print(data)
tf = Timeframe(**data, today=datetime(year=2021, month=3, day=31).date()) # note today is fixed date for easier testing
all_dates = tf.all_dates()
assert len(all_dates) == expected[0]
assert tf.n_days == expected[0]
assert tf.n_days == len(tf)
assert str(tf) == f"Timeframe: {data}"
assert tf.description == expected[1]
all_dates_expected = expected[2]
if all_dates_expected is not None:
assert all_dates == all_dates_expected
assert all_dates_expected[-1] in tf
assert not '1999-01-01' in tf
validate_date(tf.most_recent_date)
def remove_bad_stocks(df: pd.DataFrame, date_to_check: str, messages,
min_price):
"""
Remove stocks which have no data at start/end of the timeframe despite imputation that has been performed.
This ensures that profit/loss can be calculated without missing data and that optimisation is not biased towards ridiculous outcomes.
If min_price is not None, exclude all stocks with a start/end price less than min_price
"""
validate_date(date_to_check)
missing_prices = list(df.columns[df.loc[date_to_check].isna()])
if len(missing_prices) > 0:
df = df.drop(columns=missing_prices)
messages.add("Ignoring stocks with no data at {}: {}".format(
date_to_check, missing_prices))
if min_price is not None:
bad_prices = list(df.columns[df.loc[date_to_check] <= min_price])
messages.add(
f"Ignoring stocks with price < {min_price} at {date_to_check}: {bad_prices}"
)
df = df.drop(columns=bad_prices)
return df
def sum_portfolio(df: pd.DataFrame, date_str: str, stock_items):
validate_date(date_str)
portfolio_worth = sum(
map(lambda t: df.at[t[0], date_str] * t[1], stock_items))
return portfolio_worth
def test_validate_date():
validate_date('2020-01-01')
validate_date('2020-12-30')
with pytest.raises(AssertionError):
validate_date('2020-2-2')