def __init__(self, ts, windows, std_ns, is_ewm=False, is_min_periods=True):
# windows and std_ns can be either np.ndarray or single number
self.ts = np.tile(ts, (1, windows.shape[0]))
upper, middle, lower = bb_nb(ts, windows, std_ns, is_ewm,
is_min_periods)
self.upper = TimeSeries(upper)
self.middle = TimeSeries(middle)
self.lower = TimeSeries(lower)
def from_df(cls, df, labels=['Open', 'High', 'Low', 'Close', 'Volume']):
open = TimeSeries(df[labels[0]].astype(float))
high = TimeSeries(df[labels[1]].astype(float))
low = TimeSeries(df[labels[2]].astype(float))
close = TimeSeries(df[labels[3]].astype(float))
if labels[4] in df.columns:
volume = TimeSeries(df[labels[4]].astype(float))
else:
volume = None
return cls(open, high, low, close, volume)
def __init__(self,
ts,
fast_windows,
slow_windows,
is_ewm=False,
is_min_periods=True):
# fast_windows and slow_windows can be either np.ndarray or single number
fast, slow = dmac_nb(ts, fast_windows, slow_windows, is_ewm,
is_min_periods)
self.fast = TimeSeries(fast)
self.slow = TimeSeries(slow)
def drawdown(self):
return TimeSeries(1 - self.equity / expanding_max_nb(self.equity))
def returns(self):
return TimeSeries(pct_change_nb(self.equity))
def equity_in_shares(self):
return TimeSeries(self.equity / self.ts)
def equity(self):
return TimeSeries(self.cash + self.shares * self.ts)
def __init__(self, ts, windows, is_ewm=False, is_min_periods=True):
self.rsi = TimeSeries(rsi_nb(ts, windows, is_ewm, is_min_periods))
def bandwidth(self):
"""Bandwidth tells how wide the Bollinger Bands are on a normalized basis."""
return TimeSeries((self.upper - self.lower) / self.middle)
def percent_b(self):
"""Shows where price is in relation to the bands.
%b equals 1 at the upper band and 0 at the lower band."""
return TimeSeries((self.ts - self.lower) / (self.upper - self.lower))