Indicator divergence library

This module aims to help to find bullish/bearish divergences (regular or hidden) between two indicators using argrelextrema from scipy.signal.

Code based on higher-highs-lower-lows-and-calculating-price-trends-in-python article on medium.

This package is mean to use in conjunction with jesse ai but is generic enough to be used on its own.

Install

poetry build
pip install dist/jesse_extra_indicators-0.1.0-py3-none-any.whl

Example usage

import jesse_extra_indicators as xta

# Loopback window
w = 2

# this indicators has lower highs
ind1 = np.array([0, 10, 0, 9, 0]) # ex df.close
ind1_hl = xta.hl.HighLow(ind1, order=1)

# this indicators has higher highs
ind2 = np.array([0, 10, 0, 11, 0]) # ex ta.rsi(df.close)
ind2_hl = xta.hl.HighLow(ind2, order=1)

# build indicator divergence object
ind_div = xta.hl.IndicatorDivergence(ind1_hl, ind2_hl)

# Check regular or hidden bearish divergence
assert ind_div.regular_divergence(w, "bearish")
assert ind_div.hidden_divergence(w, "bearish")

# Check bearish confirmation 
assert ind_div.confirmation(w, "bearish")


# Check regular or hidden bullish divergence
assert ind_div.regular_divergence(w, "bullish")
assert ind_div.hidden_divergence(w, "bullish")

# Check bearish confirmation 
assert ind_div.confirmation(w, "bullish")

Example jesse strategy

from jesse.strategies import Strategy, cached
import jesse.indicators as ta
import jesse_extra_indicators as xta

class Example(Strategy):

    @property
    @cached
    def close_rsi_div(self):
        # no need to have the full data, this will speed up processing
        w = 100 

        close_hl = xta.hl.HighLow(self.candles[-w:, 2])
        rsi_hl = xta.hl.HighLow(self.rsi)

        # build indicator divergence object
        return xta.hl.IndicatorDivergence(close_hl, close_hl)


    def should_long(self) -> bool:
        w = 2 # you may wan to tweak the loopback window
        return (
            self.close_rsi_div.regular_divergence(w, "bullish")
            or self.close_rsi_div.hidden_divergence(w, "bullish")
        )


    def should_short(self) -> bool:
        w = 2 # you may wan to tweak the loopback window
        return (
            self.close_rsi_div.regular_divergence(w, "bearish")
            or self.close_rsi_div.hidden_divergence(w, "bearish")
        )

    def should_cancel(self) -> bool:
        ...

    def go_long(self):
        ...

    def go_short(self):
        ...

Find this usefull and want to buy me a coffee ?

send tips to

33w68oGuotfpJy59fPP5fUDk2fT3EzGkmS (btc)

D8zxqb2Fzm7Kqkn7QcKrcQjBPwbEBmMbRE (dogecoin)