def from_archive(cls, symbol, *args, **kwargs):
""" Returns an object that is capable of calculating various metrics of the stock,
with data loaded from archive.
symbol: e.g. "C6L.SI"
Any extra arguments will be passed on to the constructor.
"""
kwargs.update({
"historical": load_fundamental_data("price", symbol),
"balance_sheet": load_fundamental_data("balance_sheet", symbol),
"income_statement": load_fundamental_data("income_statement", symbol),
"cash_flow": load_fundamental_data("cash_flow", symbol),
})
return cls(*args, **kwargs)
def test_load_fundamental_data_select_columns(self):
mock_get_numerical_column_names = MagicMock(return_value=("Open", "High", "Low", "Close", "Volume", "Adj Close"))
mock_get_fundamentals = MagicMock(return_value=[
(datetime(2012, 12, 20), 6860, 100.4),
(datetime(2012, 12, 21), 7850, 99.9),
(datetime(2012, 12, 22), 3870, 32.6),
])
with patch("fa.analysis.io.get_numerical_column_names", mock_get_numerical_column_names), \
patch("fa.analysis.io.get_fundamentals", mock_get_fundamentals):
df = io.load_fundamental_data("price", "C6L.SI", ["Volume", "Adj Close"])
expected = pd.DataFrame(
[[6860, 100.4],
[7850, 99.9],
[3870, 32.6]],
index=pd.to_datetime(["2012-12-20", "2012-12-21", "2012-12-22"]),
columns=["Volume", "Adj Close"]
)
expected.index.name = "Date"
self.assertFrameEqual(df, expected)
total_gain = capital_gain + total_dividend_payout
total_return = total_gain / total_amount_invested
print("total return = {0:.0f}% after {1:.1f} years.".format(total_return * 100, cash_flow_period + 1))
dividend_payout_proportion = total_dividend_payout / total_gain
print("dividend payout comprises {0:.0f}%".format(dividend_payout_proportion * 100))
all_return = ((units_bought * sell_price + dividend_payout) / amount_invested - 1).replace(np.inf, np.nan)
annual_return = all_return / hold_period_years
print("average annual return = {0:.1f}%".format(annual_return.mean() * 100))
if __name__ == "__main__":
symbol = "^STI"
historical_data = load_fundamental_data("price", symbol)
price = historical_data["Close"]
return_in_1yr = calc_future_return(price, 365)
valuation = return_in_1yr.map(get_valuation)
valuation.index.name = "Date"
valuation.name = "Valuation"
# main working dataset for the study
dataset = pd.concat([price, valuation], axis=1).dropna()
# with open("output.csv", 'w') as f:
# dataset.to_csv(f)
# one year after the last investment
sell_date = price.index[-1]
