Workshops

QFSDK Workshop
Current location: Home > Workshops > QFSDK Workshop

Quantum Price Level Calculation (python version)

Posted:2020-06-23 16:56:03　Click:2733

Quantum Price Level Calculation (python version)

Import packages


import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import math
import datetime
import torch
from torch import nn
from torch.autograd import Variable
import matplotlib.pyplot as plt


Load Data

product = 'AUDUSD'
fileName = 'TS_data/'+'AUDUSD1440.csv'

ts = pd.read_csv(fileName,parse_dates=[0])


ts = ts.dropna(axis=0,how='any')
#print(ts.isna().any())

print(ts.shape)

DT_CL = [i for i in ts.values[:,-2]]

DT_CL



(2105, 7) 
[0.91891,
 0.92285,
 0.9298,
 0.93314,
 0.9262600000000001,
 0.92438,
 0.9322,
 0.9317799999999999,
 0.9351299999999999,
...


... 


 0.7375,
 0.7407,
 0.7370800000000001,
 ...]


Calculate Price Return 


K = []

for eL in range(0,21):
    K.append(pow((1.1924 + (33.2383*eL) + (56.2169*eL*eL))/(1 + (43.6106 *eL)),1/3))
    print("Energy Level ",eL," K",eL," = ",K[eL])


Energy Level  0  K 0  =  1.0604104260891232
Energy Level  1  K 1  =  1.2665998003610162
Energy Level  2  K 2  =  1.4911995005771048
Energy Level  3  K 3  =  1.6634993435845786
Energy Level  4  K 4  =  1.8060990099778986
Energy Level  5  K 5  =  1.929191563789632
Energy Level  6  K 6  =  2.0383229272127377
Energy Level  7  K 7  =  2.1368812897262184
Energy Level  8  K 8  =  2.2271052412903205
Energy Level  9  K 9  =  2.310559911129437
Energy Level  10  K 10  =  2.3883874710464883
Energy Level  11  K 11  =  2.4614499171710933
Energy Level  12  K 12  =  2.530415652563917
Energy Level  13  K 13  =  2.5958146503424806
Energy Level  14  K 14  =  2.658075047465853
Energy Level  15  K 15  =  2.717548251004358
Energy Level  16  K 16  =  2.7745266609153885
Energy Level  17  K 17  =  2.8292564903338833
Energy Level  18  K 18  =  2.8819472383042357
Energy Level  19  K 19  =  2.93277882031813
Energy Level  20  K 20  =  2.981907024619207


Calculate Price Return

DT_RT = []
for d in range(0, ts.shape[0]-1):
if DT_CL[d+1] > 0:
DT_RT.append(DT_CL[d]/DT_CL[d+1])
else:
DT_RT.append(1)

#  Calculate the s\mu and std dev
mu = np.mean(DT_RT)
sigma = np.std(DT_RT)

# Approximately dr
dr = (3 * sigma)/50
print(product+": mu =", mu, "; sigma = ", sigma," dr=",dr)



AUDUSD: mu = 1.0001520172318046 ; sigma =  0.00563034244868513  dr= 0.00033782054692110776





Generate wave function

auxR = 0
Q = [0 for i in range(100)]
NQ = [0 for i in range(100)]
maxRno = len(DT_RT)
tQno = 0
nQ = 0
numOfRInQSlot = 0

for r in DT_RT:
bFound = False
nQ = 0
auxR = 1 - (dr * 50) # Left boundary
while ((bFound != True) and (nQ < 100)):
if r > auxR and r <= (auxR + dr):
Q[nQ]+=1
tQno+=1
bFound = True
else:
nQ+=1
auxR = auxR + dr

# Normalize
NQ = np.array(Q)/tQno

# Find MaxQ and index of MaxQ
maxQ = NQ.max()

maxQno = NQ.tolist().index(maxQ)
print(maxQno)
print("maxQ =", maxQ, "; maxQno=", maxQno, "; total number of Q =", tQno)
r=[]
for i in range(100):
r.append(1-50*dr+i*dr)

plt.title(product+" Wavefunction")
plt.bar([i for i in range(100)], NQ, alpha=0.6,width = 0.8)



50
maxQ = 0.04331087584215592 ; maxQno= 50 ; total number of Q = 2078





Calculate Lambda L

r0  = r[maxQno] - (dr/2)
r1  = r0 + dr
rn1 = r0 - dr
Lup = (pow(rn1,2)*NQ[maxQno-1])-(pow(r1,2)*NQ[maxQno+1])
Ldw = (pow(rn1,4)*NQ[maxQno-1])-(pow(r1,4)*NQ[maxQno+1])
L   = abs(Lup/Ldw)

print(product+":", " r0 = ",r0," r1 = ",r1," r-1 = ",rn1," Q0 = ",NQ[maxQno]," Q1 = ",NQ[maxQno+1]," Q-1 = ",NQ[maxQno-1]," L = Lup/Ldw = ",Lup,"/",Ldw," = ",L)

print(product+"'s lambda = ", L)



AUDUSD:  r0 =  0.9998310897265394  r1 =  1.0001689102734606  r-1 =  0.9994932691796183  Q0 =  0.04331087584215592  Q1 =  0.03849855630413859  Q-1 =  0.0370548604427334  L = Lup/Ldw =  -0.0014942467280006297 / -0.0015447639489677908  =  0.9672977732287728
AUDUSD's lambda =  0.9672977732287728





Calculate QPR

QFEL = [0 for i in range(21)]
QPR = [0 for i in range(21)]
NQPR = [0 for i in range(21)]

# Cardano Method
for eL in range(0, 21):
p = -1 * pow((2*eL+1),2);
q = -1 * L * pow((2*eL+1),3) * pow(K[eL],3)

# Apply Cardano's Method to find the real root of the depressed cubic equation
u = pow((-0.5*q + math.sqrt(((q*q/4.0) + (p*p*p/27.0)))),1/3)
v = pow((-0.5*q - math.sqrt(((q*q/4.0) + (p*p*p/27.0)))),1/3)

QFEL[eL] = u+v

print(product+":", "Energy Level",eL," QFEL = ",QFEL[eL])

for eL in range(0, 21):
QPR[eL]  = QFEL[eL]/QFEL[0]

NQPR[eL] = 1 + 0.21*sigma*QPR[eL]



AUDUSD: Energy Level 0  QFEL =  1.3595491196477827
AUDUSD: Energy Level 1  QFEL =  4.546660186831538
AUDUSD: Energy Level 2  QFEL =  8.496416610049124
AUDUSD: Energy Level 3  QFEL =  12.928147353880194
AUDUSD: Energy Level 4  QFEL =  17.74980277460103
AUDUSD: Energy Level 5  QFEL =  22.904133519944605
AUDUSD: Energy Level 6  QFEL =  28.35133528907752
AUDUSD: Energy Level 7  QFEL =  34.061803912757995
AUDUSD: Energy Level 8  QFEL =  40.01248924070005
AUDUSD: Energy Level 9  QFEL =  46.18483576031768
AUDUSD: Energy Level 10  QFEL =  52.56352186712402
AUDUSD: Energy Level 11  QFEL =  59.13564037491598
AUDUSD: Energy Level 12  QFEL =  65.8901404508479
AUDUSD: Energy Level 13  QFEL =  72.81743314040682
AUDUSD: Energy Level 14  QFEL =  79.90910386980983
AUDUSD: Energy Level 15  QFEL =  87.1576974913583
AUDUSD: Energy Level 16  QFEL =  94.55655404371201
AUDUSD: Energy Level 17  QFEL =  102.09968090001858
AUDUSD: Energy Level 18  QFEL =  109.78165161593893
AUDUSD: Energy Level 19  QFEL =  117.597524755553
AUDUSD: Energy Level 20  QFEL =  125.54277792474099




Save QPL to file

daily_stock_data['Open']
# Initialize two-dimensional array
P_QPL = [[0 for i in range(len(daily_stock_data['Open']))] for i in range(len(NQPR))]
N_QPL = [[0 for i in range(len(daily_stock_data['Open']))] for i in range(len(NQPR))]
#Calculate +QPL
for i in range(len(NQPR)):
for j in range(len(daily_stock_data['Open'])):
P_QPL[i][j]=daily_stock_data['Open'][j]*NQPR[i]

#Calculate -QPL
for i in range(len(NQPR)):
for j in range(len(daily_stock_data['Open'])):
N_QPL[i][j]=daily_stock_data['Open'][j]/NQPR[i]

P_QPL = pd.DataFrame(P_QPL)
N_QPL = pd.DataFrame(N_QPL)
P_QPL = pd.DataFrame(P_QPL.values.T, index=P_QPL.columns, columns=P_QPL.index)
N_QPL = pd.DataFrame(N_QPL.values.T, index=N_QPL.columns, columns=N_QPL.index)
daily_stock_data=pd.concat([daily_stock_data, P_QPL], axis=1)
daily_stock_data=pd.concat([daily_stock_data, N_QPL], axis=1)
new_index = ['Date','Time','Open','High','Low','Close','Volume','QPL1','QPL2','QPL3','QPL4','QPL5','QPL6','QPL7','QPL8','QPL9','QPL10',
'QPL11','QPL12','QPL13','QPL14','QPL15','QPL16','QPL17','QPL18','QPL19','QPL20','QPL21','QPL-1','QPL-2','QPL-3','QPL-4',
'QPL-5','QPL-6','QPL-7','QPL-8','QPL-9','QPL-10','QPL-11','QPL-12','QPL-13','QPL-14','QPL-15','QPL-16','QPL-17','QPL-18','QPL-19',
'QPL-20','QPL-21']
daily_stock_data.columns = new_index
daily_stock_data.to_csv('TS_data_with_QPL/'+product+'.csv',index=False)




Copyright © 2017 Quantum Finance Forecast Center.
ICP No.1703122

到价提醒[关闭]