def indicator_table(self) -> TableData:
yahoo = YahooData()
s_list = []
for stock in self.stocks:
df = yahoo[stock]
out = self.func(df, *self.fargs, **self.fkwargs)
series = pd.Series(data=out, index=df.index, name=stock)
s_list.append(series)
df = pd.concat(s_list, axis=1)
return TableData(df)
def _convert_to_table(self, array: np.ndarray):
df = pd.DataFrame(data=array,
columns=self.stocks,
index=self.indicator_table.df.index)
return TableData(df)
STOCKS = np.array(STOCKS)
def post1(df: pd.DataFrame):
series = df[DF_ADJ_CLOSE]
ts = TrailingStats(series, 252 * 5)
stat1 = ts.exp_growth
return stat1
yahoo.symbol_names = STOCKS
indicator = Indicators(yahoo)
indicator.create(post1)
df = indicator.extract_column('post1()')
table = TableData(df)
index_spy = np.where(df.columns == 'SPY')[0][0]
# %%
class Strat1(Strategy):
def init(self):
# Build long and short rate metric
self.growths = []
# Arbitrarily set initial stock. Doesn't matter.
self.current_stocks = []
return
def next(self):
def process(df: pd.DataFrame) -> (pd.DataFrame, pd.Series):
"""Extract features and targets for ML."""
col_names = df.keys()
feature_names = [c for c in col_names if c.endswith(')')]
features = df[feature_names]
targets = df['avg_future_growth']
return features, targets
features, targets = process(df)
feature_names = features.keys()
feature_table = TableData(features)
target_table = TableData(targets)
# %% Machine learning
reg = RandomForestRegressor()
def train(features, targets, reg):
scaler_x = preprocessing.StandardScaler().fit(features.values)
# scaler_y = preprocessing.StandardScaler().fit(targets.values[:, None])
features1 = scaler_x.transform(features)
# targets1 = scaler_y.transform(targets.values[:, None]).ravel()
# x_train, x_test, y_train, y_test = train_test_split(features1,