# --- Do not remove these libs --- freqtrade backtesting --strategy SmoothScalp --timerange 20210110-20210410
from freqtrade.strategy.interface import IStrategy
from typing import Dict, List
from functools import reduce
from pandas import DataFrame
# --------------------------------
import talib.abstract as ta
import freqtrade.vendor.qtpylib.indicators as qtpylib
from typing import Dict, List
from functools import reduce
from pandas import DataFrame, DatetimeIndex, merge
# --------------------------------
import talib.abstract as ta
import freqtrade.vendor.qtpylib.indicators as qtpylib
import numpy
#V1

class heikin(IStrategy):
    INTERFACE_VERSION = 3
    #do not use this strategy in live mod. It is not good enough yet and can only be use to find trends.
    timeframe = '1h'
    #I haven't found the best roi and stoplost, so feel free to explore.
    minimal_roi = {'0': 10}
    stoploss = -0.99

    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
        dataframe['hclose'] = (dataframe['open'] + dataframe['high'] + dataframe['low'] + dataframe['close']) / 4
        dataframe['hopen'] = (dataframe['open'].shift(2) + dataframe['close'].shift(2)) / 2  #it is not the same as real heikin ashi since I found that this is better.
        dataframe['hhigh'] = dataframe[['open', 'close', 'high']].max(axis=1)
        dataframe['hlow'] = dataframe[['open', 'close', 'low']].min(axis=1)
        dataframe['emac'] = ta.SMA(dataframe['hclose'], timeperiod=6)  #to smooth out the data and thus less noise.
        dataframe['emao'] = ta.SMA(dataframe['hopen'], timeperiod=6)
        return dataframe

    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
        dataframe.loc[dataframe['emao'] < dataframe['emac'], 'entry'] = 1
        return dataframe

    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
        dataframe.loc[dataframe['emao'] > dataframe['emac'], 'exit'] = 1
        return dataframe