def download_historic_prices(trades):
# Get symbols of all stocks traded
symbols = trades.Symbol.unique().tolist()
for symbol in symbols:
symbol_trades = trades.loc[trades['Symbol'] == symbol]
start_date = symbol_trades['Date'].min().strftime("%Y-%m-%d")
# TODO: Fetch last traded date, if all sold, otherwise today's date
today = datetime.date.today().strftime("%Y-%m-%d")
# Download the historic prices from Yahoo Finance
# TODO: Handle dates when stock isn't traded on the first of the month
prices = yf.Ticker(symbol).history(start=start_date,
end=today,
interval="1mo")[['Close']]
# Drop all corporate actions (for simplicity). Later we can separate these out into
# their own files and use them to calculate portfolio performance
prices = prices.dropna()
# If historic prices exist, write them to a CSV file
if not prices.empty:
output_file = file.make_path(symbol)
prices.to_csv(output_file)
def read_trades_csv():
"""Read in the trades csv and store it in a dataframe"""
path = file.make_path('trades', '')
# Read data in from csv file
# Google Finance has some weird special character in the column name for 'Symbol'
col_names = [
'Date', 'Symbol', 'Type', 'Currency', 'Shares', 'Price', 'Commission'
]
df = pd.read_csv(path, parse_dates=['Date'], names=col_names, header=0)
# Drop rows where any column is empty
df = df.dropna(axis=0)
return df
def construct_forex_dataframe(currencies, dates):
df = pd.DataFrame(index=dates)
for currency in currencies:
if currency == BASE_CURRENCY:
df[currency] = 1
continue
path = file.make_path(currency + BASE_CURRENCY, 'forex')
dfForex = pd.read_csv(path,
index_col='Date',
names=['Date', currency],
header=0,
parse_dates=True)
df = df.join(dfForex)
return df
def read_historical_csv(symbol):
path = file.make_path(symbol)
# Read data in from csv file
try:
df = pd.read_csv(path,
index_col='Date',
parse_dates=True,
usecols=['Date', 'Close'],
na_values=['nan'])
# If there are any data errors, just print an error and skip this stock
except ValueError as err:
print('Failed to read data for: ' + symbol, err)
return None
except FileNotFoundError as err:
print(f'No pricing data for: {symbol}. Skipping this stock.')
return None
# Rename close column to symbol name
df = df.rename(columns={'Close': symbol})
return df
def download_forex(trades):
# Get the currencies of all stocks traded
currencies = trades.Currency.unique().tolist()
for currency in currencies:
# No need to fetch the base currency (no conversion needed)
if currency == BASE_CURRENCY:
continue
currency_trades = trades.loc[trades['Currency'] == currency]
start_date = currency_trades['Date'].min().strftime("%Y-%m-%d")
# TODO: Fetch last traded date if all sold, otherwise today's date
today = datetime.date.today().strftime("%Y-%m-%d")
# Download the Forex prices with the base currency from Yahoo Finance
# TODO: Handle dates when forex isn't traded on the first of the month
forex = yf.Ticker(currency + BASE_CURRENCY + '=X').history(
start=start_date, end=today, interval="1mo")[['Close']]
# If historic prices exist, write them to a CSV file
output_file = file.make_path(currency + BASE_CURRENCY, 'forex')
forex.to_csv(output_file)
def write_to_file(content, ticker_symbol):
path = file.make_path(ticker_symbol)
with open(path, 'wb') as handle:
for block in content.iter_content(1024):
handle.write(block)
def write_to_file(content, ticker):
path = file.make_path(ticker)
with open(path, 'w', newline='') as csvfile:
writer = csv.writer(csvfile)
for row in content:
writer.writerow(row)