tsferro2’s gists

tsferro2 / ml_trader.py

Created April 15, 2022 14:28

	import tpqoa
	import pickle
	import numpy as np
	import pandas as pd
	import talib as ta

	class MLTrader(tpqoa.tpqoa):
	def __init__(self, config_file, algorithm):
	super(MLTrader, self).__init__(config_file)
	self.model = algorithm['model']

tsferro2 / Optimizing DNN Hyperparameters w ParameterGrid

Created March 26, 2022 19:45

	import numpy as np
	import pandas as pd
	import talib as ta
	import pandas_datareader.data as web
	import warnings
	warnings.simplefilter('ignore')

	import random
	import tensorflow.compat.v1 as tf
	tf.disable_v2_behavior()

tsferro2 / talib-derived TA stack

Created March 26, 2022 17:46

	# MOMENTUM
	data['ADX'] = ta.ADX(data.High.values, data.Low.values,data.Close.values)
	data['ADXR'] = ta.ADXR(data.High.values, data.Low.values,data.Close.values)
	data['APO'] = ta.APO(data.Close.values)
	data['AROONdown'],data['AROONup'] = ta.AROON(data.High.values, data.Low.values)
	data['AROONdmu'] = data['AROONdown'] - data['AROONup']
	data['AROONOSC'] = ta.AROONOSC(data.High.values, data.Low.values)
	data['BOP'] = ta.BOP(data.Open.values,data.High.values, data.Low.values,data.Close.values)
	data['CCI'] = ta.CCI(data.High.values, data.Low.values,data.Close.values)
	data['CMO'] = ta.CMO(data.Close.values)

tsferro2 / Ensemble Method backtester

Created March 18, 2022 16:16

	import numpy as np
	import pandas as pd
	from pylab import plt
	import talib as ta
	from binance.client import Client
	import pandas_datareader.data as web
	import warnings
	from sklearn.naive_bayes import GaussianNB
	from sklearn.linear_model import LogisticRegression
	from sklearn.tree import DecisionTreeClassifier

tsferro2 / Jump-Diffusion time series simulation

Created March 10, 2022 17:09

	import math
	import numpy as np
	import numpy.random as npr
	from pylab import plt, mpl

	# Jump-Diffusion---------------------------------------------------------------
	# jumps/price shocks added to GBM model
	S0 = 100. # initial price
	r = 0.05 # short rate
	sigma = 0.2 # vol

tsferro2 / Stochastic Volatility-based time series simulation

Created March 10, 2022 16:50

	import math
	import numpy as np
	import numpy.random as npr
	from pylab import plt, mpl

	# Stochastic Volatility--------------------------------------------------------
	# <><><> the index is simulated using stochastic volatility <><><><><><><>
	S0 = 100. # starting index price
	r = 0.05 # short rate
	v0 = 0.25 # initial volatility

tsferro2 / Square Root Diffusion time series simulation

Created March 10, 2022 15:56

	import math
	import numpy as np
	import numpy.random as npr
	from pylab import plt, mpl

	# Square-Root Diffusion (mean-reverting process)-------------------------------
	# widely used for short rates & volatility
	x0 = 1.00 # starting value
	kappa = 3.0 # mean-reversion paramater
	theta = 1.00 # long term mean

tsferro2 / Geometric Brownian Motion time series simulation

Created March 10, 2022 15:51

	import math
	import numpy as np
	import numpy.random as npr
	from pylab import plt, mpl

	# Geometric Brownian Motion (granual/steps)
	S0 = 100 # starting price
	r = 0.05 # short rate
	sigma = 0.25 # vol
	I = 10000 # simulations

tsferro2 / Bollinger bands w MA filter backtest function

Created March 4, 2022 18:39

	def bbands_MA_backtest(data, p=(20, 2, 2, 100), plot=False, costs=True, output=True):
	data['returns'] = np.log(data['close'] / data['close'].shift(1))
	data['SMA'] = data['close'].rolling(int(p[3])).mean()

	# derive bbands data
	data['upbnd'],data['mid'],data['lwbnd'] = ta.BBANDS(data.close.values,
	p[0], p[1] , p[2]) # mid, ub, lb
	data['distance'] = data['close'] - data['mid']
	data['position'] = np.where((data['close'] > data['upbnd']) &\
	(data['close'] < data['SMA']), -1, np.nan) # shorts

tsferro2 / Bollinger bands backtest function

Created March 4, 2022 16:45

	def bbands_backtest(data, p=(20, 2, 2), plot=False, costs=True, output=True):
	data['returns'] = np.log(data['close'] / data['close'].shift(1))

	# derive bbands data
	data['upbnd'],data['mid'],data['lwbnd'] = ta.BBANDS(data.close.values,
	p[0], p[1] , p[2]) # mid, ub, lb
	data['distance'] = data['close'] - data['mid'] # center line diff
	data['position'] = np.where(data['close'] > data['upbnd'], -1, np.nan) # shorts
	data['position'] = np.where(data['close'] < data['lwbnd'], 1, data['position']) # longs
	# creates zeros which ffill below will fill up to

Tucker Sferro tsferro2