# --- Do not remove these libs ---
from freqtrade.strategy.interface import IStrategy
from pandas import DataFrame
import talib.abstract as ta
import freqtrade.vendor.qtpylib.indicators as qtpylib
# --------------------------------

class EMA520015_V17(IStrategy):
    INTERFACE_VERSION = 3
    minimal_roi = {'0': 0.15}
    # Buy and exit at market price
    order_types = {'entry': 'market', 'exit': 'market', 'stoploss': 'market', 'stoploss_on_exchange': False}
    stoploss = -0.1
    trailing_stop = True
    trailing_stop_positive = 0.01
    trailing_stop_positive_offset = 0.08
    trailing_only_offset_is_reached = True
    # Optimal timeframe for the strategy
    timeframe = '4h'

    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
        macd = ta.MACD(dataframe)
        macd = ta.MACD(dataframe, fastperiod=300, slowperiod=650, signalperiod=10)
        dataframe['macd'] = macd['macd']
        dataframe['macdhist'] = macd['macdhist']
        #Exp Moving Average (200 periods)
        dataframe['ema200'] = ta.EMA(dataframe, timeperiod=200)
        dataframe['ema5'] = ta.EMA(dataframe, timeperiod=5)
        dataframe['ema350'] = ta.EMA(dataframe, timeperiod=350)
        dataframe['ema100'] = ta.EMA(dataframe, timeperiod=100)
        dataframe['ema20'] = ta.EMA(dataframe, timeperiod=20)
        return dataframe

    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
        dataframe.loc[(dataframe['close'].shift(1) < dataframe['ema20']) & (dataframe['close'] > dataframe['ema20']), 'entry'] = 1
        return dataframe

    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
        dataframe.loc[(dataframe['close'].shift(1) > dataframe['ema20']) & (dataframe['close'] < dataframe['ema20']), 'exit'] = 1
        return dataframe

    def custom_exit(self, pair: str, trade: 'Trade', current_time: 'datetime', current_rate: float, current_profit: float, **kwargs):
        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
        last_candle = dataframe.iloc[-1].squeeze()
        if current_profit > 0.02 and last_candle['ema20'] < last_candle['ema200']:
            return 'exit2'