-
Notifications
You must be signed in to change notification settings - Fork 0
/
main_bashe.py
165 lines (134 loc) · 6.53 KB
/
main_bashe.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
from minizinc import Solver, Model, Instance
import numpy as np
import pandas as pd
import datetime
import matplotlib.pyplot as plt
import streamlit as st
gold = pd.read_csv("GLD.csv")
stock = pd.read_csv("^IXIC.csv")
budget = 50000
stockAmount = 0.0
goldAmount = 0
bondAmount = []
startTime = datetime.datetime.strptime("2023-05-01", "%Y-%m-%d")
endTime = datetime.datetime.strptime("2023-11-29", "%Y-%m-%d")
placeholder = st.empty()
# functions
def saleBond(value):
for d in range(1, 28):
for bond in bondAmount:
if value == 0:
break
if (startTime - bond[1]).days % 28 == d and (startTime - bond[1]).days > 28:
if bond[0] > value:
bond[0] -= value
value = 0
break
else:
value -= bond[0]
bond[0] = 0
if value <= 0:
break
def dateToNumber(dates):
datesInNumber = []
for d in dates:
tempTime = datetime.datetime.strptime(d, "%Y-%m-%d")
datesInNumber.append((tempTime - startTime).days)
return datesInNumber
def getLastMonthData(decisionDate, dataFrame):
dates = (dataFrame[(dataFrame["Date"] <= decisionDate.strftime("%Y-%m-%d")) & (dataFrame["Date"]
>= (decisionDate - datetime.timedelta(days=4)).strftime("%Y-%m-%d"))]["Date"].to_list())
prices = (dataFrame[(dataFrame["Date"] <= decisionDate.strftime("%Y-%m-%d")) & (dataFrame["Date"]
>= (decisionDate - datetime.timedelta(days=4)).strftime("%Y-%m-%d"))]["Close"].to_list())
return dates, prices
# functions end
while startTime <= endTime:
# gold close price
dates, goldClosedPrice = getLastMonthData(startTime, gold)
# stock close price
dates, stockClosedPrice = getLastMonthData(startTime, stock)
solver = Solver.lookup("cbc")
model = Model("./Prediction.mzn")
instance_gold = Instance(solver, model)
instance_gold["numberOfDays"] = len(dates)
instance_gold["days"] = [float(i) for i in dateToNumber(dates)]
instance_gold["price"] = goldClosedPrice
result_gold = instance_gold.solve()
instance_stock = Instance(solver, model)
instance_stock["numberOfDays"] = len(dates)
instance_stock["days"] = [float(i) for i in dateToNumber(dates)]
instance_stock["price"] = stockClosedPrice
result_stock = instance_stock.solve()
# plotting
# plt.plot(dates, goldClosedPrice, dates, [
# float(result_gold["a"]) * d + float(result_gold["b"]) for d in dateToNumber(dates)])
# plt.show()
# plt.plot(dates, stockClosedPrice, dates, [
# float(result_stock["a"]) * d + float(result_stock["b"]) for d in dateToNumber(dates)])
# plt.show()
# Convert dates to numbers
date_numbers = dateToNumber(dates)
# Set the width of the figure
# Adjust the width (10 inches) and height (6 inches) as needed
fig, axs = plt.subplots(2, 1, figsize=(10, 6))
# 1st subplot
axs[0].plot(dates, goldClosedPrice, label='Gold Closed Price')
axs[0].plot(dates, [float(result_gold["a"]) * d + float(result_gold["b"])
for d in date_numbers], label='Regression Line')
axs[0].legend()
# 2nd subplot
axs[1].plot(dates, stockClosedPrice, label='Stock Closed Price')
axs[1].plot(dates, [float(result_stock["a"]) * d + float(result_stock["b"])
for d in date_numbers], label='Regression Line')
axs[1].legend()
# # Adjust layout to prevent clipping of labels
plt.tight_layout()
# Show the plot
# plt.show()
model_decision = Model("./main.mzn")
instance_decision = Instance(solver, model_decision)
instance_decision["budget"] = budget
instance_decision["goldAmount"] = goldAmount
instance_decision["stockAmount"] = stockAmount
instance_decision["saleableBondAmount"] = sum([
i[0] for i in bondAmount if i[1] < startTime - datetime.timedelta(weeks=4)
])
instance_decision["bondAmount"] = sum([
i[0] for i in bondAmount
])
instance_decision["goldCurrentPrice"] = goldClosedPrice[-1]
instance_decision["goldPredictedPrice"] = result_gold["a"] * \
(dateToNumber(dates)[-1]+7) + result_gold["b"]
instance_decision["stockCurrentPrice"] = stockClosedPrice[-1]
instance_decision["stockPredictedPrice"] = result_stock["a"] * \
(dateToNumber(dates)[-1]+7) + result_stock["b"]
result_decision = instance_decision.solve()
print(result_decision)
goldAmount += result_decision["goldDiff"]
stockAmount += result_decision["stockDiff"]
if result_decision["bondDiff"] > 0:
bondAmount.append([result_decision["bondDiff"], startTime])
elif result_decision["bondDiff"] < 0:
saleBond(abs(result_decision["bondDiff"]))
budget = budget - (result_decision["goldDiff"] * instance_decision["goldCurrentPrice"] +
result_decision["stockDiff"] * instance_decision["stockCurrentPrice"] + result_decision["bondDiff"])
for i in bondAmount:
if (startTime - i[1]).days % 28 == 0 and (i[1] - startTime).days != 0:
budget += i[0] * 0.0054
print(
f"date:{startTime.strftime('%Y-%m-%d')}, total : {budget + goldAmount * goldClosedPrice[-1] + stockAmount * stockClosedPrice[-1] + sum([i[0] for i in bondAmount])}, budget: {budget}, goldAmount: {goldAmount}, stockAmount: {stockAmount}, bondAmount: {sum([i[0] for i in bondAmount])}")
print()
with placeholder.container():
st.title(startTime.strftime('%Y-%m-%d'))
st.metric(label="Total", value=budget, delta=goldAmount *
goldClosedPrice[-1] + stockAmount * stockClosedPrice[-1] + sum([i[0] for i in bondAmount]))
st.metric(label="Gold Amount: ", value=goldAmount,
delta=result_decision["goldDiff"])
st.metric(label="Stock Amount: ", value=stockAmount,
delta=result_decision["stockDiff"])
st.metric(label="Bond Amount: ", value=sum(
[i[0] for i in bondAmount]), delta=result_decision["bondDiff"])
st.metric(label="Total Budget: ", value=budget + goldAmount *
goldClosedPrice[-1] + stockAmount * stockClosedPrice[-1] + sum([i[0] for i in bondAmount]), delta=result_decision["goldDiff"] * goldClosedPrice[-1] + result_decision["stockDiff"] * stockClosedPrice[-1] + result_decision["bondDiff"])
st.pyplot(fig)
startTime += datetime.timedelta(weeks=1)