Use of talib indicators not producing the expected result.

[selvanponraj]

TA lib indicators not producing expected results. Any help on this would be great.

self.ma1 = self.I(SMA, Close, 10)  # Return type is _indicator

self.ma1 = ta.SMA(self.data.Close, timeperiod=10)  # Return type is ndarray.

And also I am using this function and getting the error

        if qtpylib.crossed_above(self.ma1,self.ma2):
            self.buy()
        elif qtpylib.crossed_above(self.ma2,self.ma1):
            self.sell()

File "/Users/vt_qtpylib/lib/python3.7/site-packages/pandas/core/generic.py", line 1479, in __nonzero__
    f"The truth value of a {type(self).__name__} is ambiguous. "
ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

Expected Behavior

self.ma1 = self.I(SMA, Close, 10)
self.ma2 = self.I(SMA, Close, 20)

if crossover(self.ma11,self.ma22):
        self.buy()
elif crossover(self.ma22,self.ma11):
        self.sell()

Start                     2020-04-01 23:00:00
End                       2020-04-30 23:59:00
Duration                     29 days 00:59:00
Exposure [%]                          99.8733
Equity Final [$]                      744.094
Equity Peak [$]                         10000
Return [%]                           -92.5591
Buy & Hold Return [%]                 15.9654
Max. Drawdown [%]                    -92.5934
Avg. Drawdown [%]                    -92.5934
Max. Drawdown Duration       29 days 00:24:00
Avg. Drawdown Duration       29 days 00:24:00
# Trades                                 1155
Win Rate [%]                          13.2468
Best Trade [%]                        2.14186
Worst Trade [%]                      -1.69105
Avg. Trade [%]                      -0.224556
Max. Trade Duration           3 days 01:17:00
Avg. Trade Duration           0 days 00:37:00
Expectancy [%]                       0.305656
SQN                                  -17.5662
Sharpe Ratio                        -0.776904
Sortino Ratio                        -1.29619
Calmar Ratio                      -0.00242518
_strategy                            SmaCross

Actual Behavior

Use of TALib

self.ma1 = ta.SMA(self.data.Close, timeperiod=10)
self.ma2 = ta.SMA(self.data.Close, timeperiod=20)

if crossover(self.ma11,self.ma22):
        self.buy()
elif crossover(self.ma22,self.ma11):
        self.sell()

TALib Produces NAN values.

Start                     2020-04-01 23:00:00
End                       2020-04-30 23:59:00
Duration                     29 days 00:59:00
Exposure [%]                                0
Equity Final [$]                        10000
Equity Peak [$]                         10000
Return [%]                                  0
Buy & Hold Return [%]                 15.9654
Max. Drawdown [%]                          -0
Avg. Drawdown [%]                         NaN
Max. Drawdown Duration                    NaN
Avg. Drawdown Duration                    NaN
# Trades                                    0
Win Rate [%]                              NaN
Best Trade [%]                            NaN
Worst Trade [%]                           NaN
Avg. Trade [%]                            NaN
Max. Trade Duration                       NaT
Avg. Trade Duration                       NaT
Expectancy [%]                            NaN
SQN                                       NaN
Sharpe Ratio                              NaN
Sortino Ratio                             NaN
Calmar Ratio                              NaN
_strategy                            SmaCross

[kernc]

File "/Users/vt_qtpylib/lib/python3.7/site-packages/pandas/core/generic.py", line 1479, in __nonzero__
    f"The truth value of a {type(self).__name__} is ambiguous. "
ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

Can you provide a full traceback?

Can you actually provide a full MWE to reproduce the issue?

Strategy.I returning _Indicator instead of np.ndarray should not be an issue as the former is a subtype of the latter:

backtesting.py/backtesting/_util.py

Line 43 in 64bcda2

class _Array(np.ndarray):

backtesting.py/backtesting/_util.py

Line 88 in 64bcda2

class _Indicator(_Array):

[kernc]

Actual Behavior

Use of TALib

self.ma1 = ta.SMA(self.data.Close, timeperiod=10)

The only proper way to use talib (or any indicator functions, really) is:

self.ma1 = self.I(ta.SMA, self.data.Close, timeperiod=10)

I.e. lazy evaluation. Just pass the function reference (ta.SMA) to Strategy.I and let it call it.

[selvanponraj]

Thanks for your help.

Please find the MWE.

from backtesting import Backtest, Strategy
from backtesting.lib import *
from backtesting.test import *
from talib import abstract as ta

class SmaCross(Strategy):

    def init(self):
        # Precompute the two moving averages
        self.sma1 = self.I(SMA, self.data.Close, 10)
        self.sma2 = self.I(SMA, self.data.Close, 20)

        # self.sma1 = self.I(ta.SMA, self.data.Close, timeperiod=10)
        # self.sma2 = self.I(ta.SMA, self.data.Close, timeperiod=20)

    def next(self):
        # If sma1 crosses above sma2, close any existing
        # short trades, and buy the asset
        if crossover(self.sma1, self.sma2):
            self.position.close()
            self.buy()

        # Else, if sma1 crosses below sma2, close any existing
        # long trades, and sell the asset
        elif crossover(self.sma2, self.sma1):
            self.position.close()
            self.sell()

bt = Backtest(GOOG, SmaCross,
              cash=10000, commission=.002)

print(bt.run())
# bt.plot()

Same code using TA Lib throws exception

self.sma1 = self.I(ta.SMA, self.data.Close, timeperiod=10)
self.sma2 = self.I(ta.SMA, self.data.Close, timeperiod=20)

Traceback (most recent call last):
  File "/local/backtesting.py/doc/examples/SmaCross.py", line 33, in <module>
    print(bt.run())
  File "/local/backtesting.py/backtesting/backtesting.py", line 1073, in run
    strategy.init()
  File "/local/backtesting.py/doc/examples/SmaCross.py", line 14, in init
    self.sma1 = self.I(ta.SMA, self.data.Close, timeperiod=10)
  File "/local/backtesting.py/backtesting/backtesting.py", line 122, in I
    func_name = _as_str(func)
  File "/local/backtesting.py/backtesting/_util.py", line 25, in _as_str
    name = value.__name__.replace('<lambda>', 'λ')
AttributeError: 'Function' object has no attribute '__name__'

[kernc]

  File "/local/backtesting.py/doc/examples/SmaCross.py", line 14, in init
    self.sma1 = self.I(ta.SMA, self.data.Close, timeperiod=10)
  File "/local/backtesting.py/backtesting/backtesting.py", line 122, in I
    func_name = _as_str(func)
  File "/local/backtesting.py/backtesting/_util.py", line 25, in _as_str
    name = value.__name__.replace('<lambda>', 'λ')
AttributeError: 'Function' object has no attribute '__name__'

This is now a duplicate of issue #93, and a different error from the one quoted in original post? 😕

File "/Users/vt_qtpylib/lib/python3.7/site-packages/pandas/core/generic.py", line 1479, in __nonzero__
    f"The truth value of a {type(self).__name__} is ambiguous. "
ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

The workaround, besides using yet-unreleased git master, is to instead:

import talib as ta

[selvanponraj]

Many thanks and workaround works.