Non used indicator changes results.

[Nireus79]

def init(self):
    close = self.data.Close
    self.L_atr = self.I(atr_wrapper, self.data.df)
    self.M_K = resample_apply('1H', k_wrapper, self.data.df)
    self.M_D = self.I(sma_indicator, pd.Series(self.M_K), 3, fillna=False)
    self.H_ema1 = resample_apply('6H', ema_indicator, close, self.ema1, fillna=False)
    self.H_ema2 = resample_apply('6H', ema_indicator, close, self.ema2, fillna=False)

def next(self):
    close = self.data.Close
    high = self.data.High
    low = self.data.Low
    self.trend1W = close > self.H_ema1
    self.buy_flag1D = self.M_K > self.M_D > self.lower_bound
    self.sell_flag1D = self.M_K < self.M_D < self.upper_bound

I set my indicators for different timeframes under init. Then I do not use the second ema.
The problem is that when I delete or comment the second ema from init section I get different results.
It is an indicator I do not call anywhere within my code and still changes results only because it is calculated.
ema2 is also giving different results when optimized.
What do I do wrong? I cannot figure it out. Please guide me.
Thank you.

[kernc]

See #674 or #732. Supposedly fixed in d7eaa45, to be released.

[Nireus79]

Thank you very much.