# source: https://raw.githubusercontent.com/ivanproskuryakov/ta-search/6d0c9b6c558c0c0ec7c515b96c736ebc589bf78a/user_data/strategies/TaSearchLevelJ15m.py
import pandas as pd
import numpy as np
import talib.abstract as ta
import logging

from datetime import datetime
from typing import Dict, List, Optional, Tuple, Union
from scipy import signal
from statistics import mean

from freqtrade.persistence.trade_model import Trade
from freqtrade.strategy.interface import IStrategy

# Main
# "minimal_roi": {
#     "0": 0.05
# },
# "stoploss": -0.05,
#
# "trailing_stop": true,
# "trailing_stop_positive": 0.05,
# "trailing_stop_positive_offset": 0.2,
# "trailing_only_offset_is_reached": true,

class github_ivanproskuryakov_ta_search__TaSearchLevelJ15m__20230412_084120(IStrategy):
    can_short: bool = True

    def populate_indicators(self, df: pd.DataFrame, metadata: dict) -> pd.DataFrame:
        pd.set_option('display.max_rows', 100000)
        pd.set_option('display.precision', 10)
        pd.set_option('mode.chained_assignment', None)

        df['rsi_7'] = ta.RSI(df['close'], timeperiod=7).round(2)
        df['min_level'] = 0
        df['max_level'] = 0
        df['buy_short'] = 0
        df['buy_long'] = 0
        df['buy_short2'] = 0
        df['buy_long2'] = 0
        df['i_close'] = 0
        df['i_open'] = 0
        df['i_low'] = 0
        df['i_high'] = 0

        logging.getLogger('freqtrade').info(str(metadata))

        df = self.do_heikin_ashi(df)
        df = self.do_long(df)
        df = self.do_short(df)

        return df

    def do_long(self, df: pd.DataFrame) -> pd.DataFrame:
        n = 200
        df['min_local'] = df.iloc[signal.argrelextrema(df.c.values, np.less_equal, order=n)[0]]['c']
        min = df['c'].max()

        times = df.query(f'min_local > 0')
        prices = []

        if len(times) > 0:
            for i in range(500, len(df)):
                if df['min_local'].loc[i] > 0:
                    close = df['c'].loc[i]

                    chunk = df[:i - 10]
                    chunk['min_local'] = chunk.iloc[signal.argrelextrema(chunk.c.values, np.less_equal, order=n)[0]]['c']

                    time_chunk = chunk.query(f'min_local > 0')

                    for x, row in time_chunk.iterrows():
                        close_chunk = time_chunk['c'].loc[x]
                        prices.append(close_chunk)

                    for p in prices:
                        diff = self.diff_percentage(p, close)

                        if 0 < diff < 0.5:
                            logging.getLogger('freqtrade').info(str([i, '+++', diff, p]))
                            df['min_level'].loc[i] = 1

                x0 = i - 100
                for x in range(i - 100, i):
                    if df['min_level'].loc[x] > 0:
                        diff = self.diff_percentage(df['c'].loc[x], df['c'].loc[i])
                        xt = df.query(f'buy_short > 0 and {x0} < index < {i}')

                        if 0 < diff < 0.5:
                            df['buy_long'].loc[i] = 1

                            if len(xt) > 1:
                                #print(len(xt))
                                logging.getLogger('freqtrade').info(
                                    str([i, '---- long2 ----'])
                                )
                                df['buy_long2'].loc[i] = 1
                                df['i_low'].loc[i] = df['low'].loc[x]
                                df['i_high'].loc[i] = df['high'].loc[x]
                                df['i_open'].loc[i] = df['o'].loc[x]
                                df['i_close'].loc[i] = df['c'].loc[x]

        return df

    def do_short(self, df: pd.DataFrame) -> pd.DataFrame:
        n = 200
        df['max_local'] = df.iloc[signal.argrelextrema(df.c.values, np.greater_equal, order=n)[0]]['c']
        max = df['c'].min()

        times = df.query(f'max_local > 0')
        prices = []

        if len(times) > 0:
            for i in range(500, len(df)):
                if df['max_local'].loc[i] > 0:
                    close = df['c'].loc[i]

                    chunk = df[:i - 10]
                    chunk['max_x'] = chunk.iloc[signal.argrelextrema(chunk.c.values, np.greater_equal, order=n)[0]]['c']

                    time_chunk = chunk.query(f'max_x > 0')

                    for x, row in time_chunk.iterrows():
                        close_chunk = time_chunk['c'].loc[x]
                        prices.append(close_chunk)

                    for p in prices:
                        diff = self.diff_percentage(p, close)

                        if 0 < diff < 0.5:
                            logging.getLogger('freqtrade').info(str([i, '---', diff, p]))
                            df['max_level'].loc[i] = 1

                x0 = i - 100
                for x in range(x0, i):
                    if df['max_level'].loc[x] > 0:
                        diff = self.diff_percentage(df['c'].loc[x], df['c'].loc[i])
                        xt = df.query(f'buy_short > 0 and {x0} < index < {i}')

                        if 0 < diff < 0.5:
                            df['buy_short'].loc[i] = 1

                            if len(xt) > 1:
                                #print(len(xt))
                                logging.getLogger('freqtrade').info(
                                    str([i, '---- short2 ----'])
                                )
                                df['buy_short2'].loc[i] = 1
                                df['i_low'].loc[i] = df['low'].loc[x]
                                df['i_high'].loc[i] = df['high'].loc[x]
                                df['i_open'].loc[i] = df['o'].loc[x]
                                df['i_close'].loc[i] = df['c'].loc[x]


        return df

    def populate_entry_trend(self, df: pd.DataFrame, metadata: dict) -> pd.DataFrame:
        # df.loc[(df['buy_short'] > 0), 'enter_short'] = 1
        df.loc[(df['buy_short2'] > 0), 'enter_short'] = 1

        # df.loc[(df['buy_long'] > 0), 'enter_long'] = 1
        df.loc[(df['buy_long2'] > 0), 'enter_long'] = 1

        return df

    def populate_exit_trend(self, df: pd.DataFrame, metadata: dict) -> pd.DataFrame:
        df.loc[
            (df['rsi_7'] < 10),
            'exit_short'
        ] = 1
        df.loc[
            (df['rsi_7'] > 90),
            'exit_long'
        ] = 1

        return df

    def leverage(self, pair: str, current_time: datetime, current_rate: float,
                 proposed_leverage: float, max_leverage: float, entry_tag: Optional[str],
                 side: str, **kwargs) -> float:

        return 10

    def diff_percentage(self, v2, v1) -> float:
        diff = ((v2 - v1) / ((v2 + v1) / 2)) * 100
        diff = np.round(diff, 4)

        return np.abs(diff)

    def do_heikin_ashi(self, df: pd.DataFrame) -> pd.DataFrame:
        """
        https://www.investopedia.com/trading/heikin-ashi-better-candlestick/
        """
        df['h'] = df.apply(lambda x: max(x['high'], x['open'], x['close']), axis=1)
        df['l'] = df.apply(lambda x: min(x['low'], x['open'], x['close']), axis=1)

        for i in range(1, len(df)):
            df.loc[i, 'c'] = 1 / 4 * (
                df['open'].iloc[i] +
                df['close'].iloc[i] +
                df['high'].iloc[i] +
                df['low'].iloc[i]
            )
            df.loc[i, 'o'] = 1 / 2 * (
                df['open'].iloc[i - 1] +
                df['close'].iloc[i - 1]
            )

        return df