example_12345_impulsive_wave.py

from __future__ import annotations
from models.WavePattern import WavePattern
from models.WaveRules import Impulse, LeadingDiagonal
from models.WaveAnalyzer import WaveAnalyzer
from models.WaveOptions import WaveOptionsGenerator5
from models.helpers import plot_pattern
import pandas as pd
import numpy as np
import yfinance as yf
from pprint import pprint

# end_date = pd.Timestamp.now()
# start_date = end_date - pd.DateOffset(days=360)
# df = yf.download('MSFT', start=start_date, end=end_date)

df = pd.read_csv(r'data\btc-usd_1d.csv')
idx_start = np.argmin(np.array(list(df['Low'])))

wa = WaveAnalyzer(df=df, verbose=False) # .reset_index()
wave_options_impulse = WaveOptionsGenerator5(up_to=15)  # generates WaveOptions up to [15, 15, 15, 15, 15]

impulse = Impulse('impulse')
leading_diagonal = LeadingDiagonal('leading diagonal')
rules_to_check = [impulse, leading_diagonal]

print(f'Start at idx: {idx_start}')
print(f"will run up to {wave_options_impulse.number / 1e6}M combinations.")

# set up a set to store already found wave counts
# it can be the case, that 2 WaveOptions lead to the same WavePattern.
# This can be seen in a chart, where for example we try to skip more maxima as there are. In such a case
# e.g. [1,2,3,4,5] and [1,2,3,4,10] will lead to the same WavePattern (has same sub-wave structure, same begin / end,
# same high / low etc.
# If we find the same WavePattern, we skip and do not plot it

wavepatterns_up = set()

# loop over all combinations of wave options [i,j,k,l,m] for impulsive waves sorted from small, e.g.  [0,1,...] to
# large e.g. [3,2, ...]
for new_option_impulse in wave_options_impulse.options_sorted:

    waves_up = wa.find_impulsive_wave(idx_start=idx_start, wave_config=new_option_impulse.values)

    if waves_up:
        wavepattern_up = WavePattern(waves_up, verbose=True)

        for rule in rules_to_check:

            if wavepattern_up.check_rule(rule):
                if wavepattern_up in wavepatterns_up:
                    continue
                else:
                    wavepatterns_up.add(wavepattern_up)
                    print(f'{rule.name} found: {new_option_impulse.values}')
                    plot_pattern(df=df, wave_pattern=wavepattern_up, title=str(new_option_impulse))