feature_functions.py

import pandas as pd
import numpy as np
from scipy import stats
import scipy.optimize
from scipy.optimize import OptimizeWarning
import warnings
import math
import matplotlib.pyplot as plt
from sklearn.linear_model import LinearRegression
from matplotlib.finance import _candlestick
from matplotlib.dates import date2num
from datetime import datetime

class holder:


# Heiken Ashi Candles

def heikenashi(prices.periods):

    """
    :param prices: dataframe of OHLC & volume data
    :param periods: periods for which to create the candles
    :return: heiken ashi OHLC candles
    """

    results = holder()

    # store candles in a dataframe in a dict
    dict = {}

    HAclose = prices['open', 'high', 'low', 'close'].sum(axis=1)/4
    HAopen = HAclose.copy()
    HAopen.iloc[0] = HAclose.iloc[0]
    HAhigh = HAclose.copy()
    HAlow = HAclose.copy()

    for i in range(1,len(prices)):
        HAopen.iloc[i] = (HAopen.iloc[i-1]+HAclose.iloc[i-1])/2
        HAhigh.iloc[i] = np.array([prices.high.iloc[i],HAopen.iloc[i],HAclose.iloc[i]]).max()
        HAlow.iloc[i] = np.array([prices.low.iloc[i], HAopen.iloc[i], HAclose.iloc[i]]).min()

    df = pd.concat((HAopen,HAhigh,HAlow,HAclose),axis=1)
    df.columns = ['open', 'high', 'low', 'close']

    #df.index = df.index.droplevel(0)

    dict[periods[0]] = df

    results.candles = dict

    return results