Yuttakan Sangchai ProAek11

AI-Frendersis - Strategy Trading

ProAek11 / MQL5_GetFinCnadle.py

Created May 20, 2020 15:43 — forked from igotit-anything/MQL5_GetFinCnadle.py

Oython easy get financial data from MetaTrader5

	from MetaTrader5 import *
	import pandas as pd
	pd.set_option('display.max_columns', 500) # number of columns to be displayed
	pd.set_option('display.width', 1500) # max table width to display
	import matplotlib.pyplot as plt

	MT5Initialize() # connect to MT5
	MT5WaitForTerminal() # wait till MT5 ready

	### get candle data

ProAek11 / simulator.py

Created April 24, 2020 11:51 — forked from markusrenepae/simulator.py

This gist is for another medium article and is about an investment simulator.

	import pandas as pd
	import numpy as np
	import datetime as dt
	import math
	import warnings

	warnings.filterwarnings("ignore")

	prices = pd.read_csv("adjclose.csv", index_col="Date", parse_dates=True)
	volumechanges = pd.read_csv("volume.csv", index_col="Date", parse_dates=True).pct_change()*100

ProAek11 / pytorch_data.py

Created March 10, 2020 18:18 — forked from witchapong/pytorch_data.py

	class StockDataset(Dataset):
	def __init__(self, df_seq, feat_num, seq_len, target_len, df_cat):
	# SEQUENTIAL PART
	self.df_seq = df_seq.iloc[:,:-target_len]
	self.df_cat = df_cat
	self.target = df_seq.iloc[:,-target_len:]

	def __getitem__(self, index):
	return(torch.tensor(self.df_seq.iloc[index].values.reshape(seq_len,feat_num), dtype=torch.float, device=device),
	torch.tensor(self.df_cat.iloc[index], dtype=torch.long, device=device),

ProAek11 / hmm_market_behavior_run_algorithm_catalyst.py

Created March 9, 2020 11:28 — forked from lamres/hmm_market_behavior_run_algorithm_catalyst.py

ProAek11 / hmm_market_behavior_analyze_catalyst.py

Created March 9, 2020 11:28 — forked from lamres/hmm_market_behavior_analyze_catalyst.py

	def analyze(context, perf):
	sns.set()

	# Summary output
	print("Total return: " + str(perf.algorithm_period_return[-1]))
	print("Sortino coef: " + str(perf.sortino[-1]))
	print("Max drawdown: " + str(np.min(perf.max_drawdown[-1])))
	print("alpha: " + str(perf.alpha[-1]))
	print("beta: " + str(perf.beta[-1]))

ProAek11 / hmm_market_behavior_handle_data_catalyst.py

Created March 9, 2020 11:27 — forked from lamres/hmm_market_behavior_handle_data_catalyst.py

	def handle_data(context, data):
	current_date = get_datetime().date()
	current_time = get_datetime().time()

	# Just one time in a day (first minute)
	if current_time.hour == 0 and current_time.minute == 0 and current_time.second == 0:
	prices = pd.DataFrame()
	volumes = pd.DataFrame()

	try:

ProAek11 / hmm_market_behavior_initialize_catalyst.py

Created March 9, 2020 11:27 — forked from lamres/hmm_market_behavior_initialize_catalyst.py

	def initialize(context):
	context.asset = symbol('btc_usd')
	context.leverage = 1.0

	context.std_period = 10
	context.ma_period = 10
	context.price_deviation_period = 10
	context.volume_deviation_period = 10
	context.n_periods = 5 + int(np.max([context.std_period, context.ma_period,
	context.price_deviation_period, context.volume_deviation_period]))

ProAek11 / hmm_market_behavior_libraries_catalyst.py

Created March 9, 2020 11:27 — forked from lamres/hmm_market_behavior_libraries_catalyst.py

	import matplotlib.pyplot as plt
	import numpy as np
	import pandas as pd
	import joblib
	from hmmlearn.hmm import GaussianHMM
	import datetime
	import seaborn as sns

	from catalyst import run_algorithm
	from catalyst.api import (record, symbol, order_target_percent, date_rules, time_rules, get_datetime)

ProAek11 / hmm_market_behavior_save_model.py

Created March 9, 2020 11:26 — forked from lamres/hmm_market_behavior_save_model.py

joblib.dump(model, 'quandl_' + asset.replace('/', '_') + '_final_model.pkl')

ProAek11 / hmm_market_behavior_compare_hidden_states.py

Created March 9, 2020 11:25 — forked from lamres/hmm_market_behavior_compare_hidden_states.py

	def mean_confidence_interval(vals, confidence):
	a = 1.0 * np.array(vals)
	n = len(a)
	m, se = np.mean(a), scipy.stats.sem(a)
	h = se * scipy.stats.t.ppf((1 + confidence) / 2., n-1)
	return m - h, m, m + h

	def compare_hidden_states(hmm_model, cols_features, conf_interval, iters = 1000):
	plt.figure(figsize=(15, 15))
	fig, axs = plt.subplots(len(cols_features), hmm_model.n_components, figsize = (15, 15))