import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from matplotlib import interactive, cm
from time import time
#from math import log10
interactive(True)


def dataset(nr):
    if (nr==1): 
        name = "AAPL-84-19.csv"
        c = 0
    elif(nr==2):
        name = "INTC-84-19.csv"
        c = 1900
    elif(nr==3):
        name = "IXIC-84-19.csv"
        c = 1900
    x = pd.read_csv(name)
    print('dataset: ',name)
    
# preview the first and last lines of data
    print(x.head(2))
    print(x.tail(2))
    print('type & size of x ',type(x),np.size(x))
    y = x.as_matrix()
    print('type & size of y ',type(y),np.size(y))
    print('y[0,...]',y[0,:] )
    Nd = np.size(y)//7 
    print('Nd =',Nd)
    t = np.zeros(Nd,dtype=int)
    ymd = np.ndarray((Nd,3),dtype=int)
    po = y[:,1]*100;    pc = y[:,4]*100
    ph = y[:,2]*100;    vol = y[:,6]
    pl = y[:,3]*100

    for n in range(Nd):
    # decode dates, store numeric values in ymd array
        date = y[n,0]
        ymd[n,0] = yy = int(np.array(date.split('-')[0])) - c
        ymd[n,1] = mm = int(np.array(date.split('-')[1]))
        ymd[n,2] = dd = int(np.array(date.split('-')[2]))
        t[n] = n
    print('n=0 (beg.):', ymd[0,0], ymd[0,1], ymd[0,2],'\n')
    return(Nd,y,ymd)


'''
   read and process AAPL, INTC, and IXIC historical data
   in .csv format, 1984-2019.
'''
# AAPL
NdA,yA,ymdA = dataset(1)
# INTC
NdI,yI,ymdI = dataset(2)
# IXIC = NASDAQ
NdX,yX,ymdX = dataset(3)

N = np.size(yA)//7  # of rows, numbered from 0 to N-1
print('N=',N)
pA = np.array(yA[:,4],dtype=float)  # without explicit dtype=float there will be 
pI = np.array(yI[:,4],dtype=float)  # problems with math functions like np.log10(pA)
pX = np.array(yX[:,4],dtype=float)

print('logA',type(pA),type(np.array([1,3.3])),np.log10(np.array([1,3.3])))
A = (pA/pA[0])
I = (pI/pI[0])
X = (pX/pX[0])
print(np.size(A),np.size(ymdA))
print(np.size(I),np.size(ymdI))
print(np.size(X),np.size(ymdX))


t = np.linspace(-N/240,0,N) 
# take a look at the data
plt.semilogy(t,A,color=(1,0,0))
plt.semilogy(t,I,color=(0,0,1))
plt.semilogy(t,X,color=(0,0,0))
plt.title('INTC, IXIC=NASDAQ, AAPL')
plt.xlabel(' time [yr]')
plt.ylabel(' price normalizd to initial price')
plt.grid()
plt.show()
input(' ok?')
#
plt.cla()
plt.semilogy(t,A/X,color=(1,0,0))
plt.semilogy(t,I/X,color=(0,0,1))
plt.semilogy(t,X/X,color=(0,0,0))
plt.title('INTC/IXIC, AAPL/IXIC')
plt.xlabel(' time [yr]')
plt.ylabel(' price relative to market average')
plt.grid()
plt.show()
input(' ok?')

# a bit of analysis 
lA = np.log10(A)
# (...)