def gen_tree(q, entry):
u = u_and_d.u(q, entry)
d = u_and_d.d(q, entry)
day_0 = Node(0, 0, fetch.close_data(entry)[-5], 1)
day_1_u = Node(1, 1, day_0.price * u, q)
day_1_d = Node(2, 1, day_0.price * d, (1 - q))
day_2_uu = Node(3, 2, day_1_u.price * u, q * q)
day_2_ud = Node(4, 2, day_1_u.price * d, 2 * q * (1 - q))
day_2_dd = Node(5, 2, day_1_d.price * d, (1 - q) * (1 - q))
day_3_uuu = Node(6, 3, day_2_uu.price * u, q * q * q)
day_3_uud = Node(7, 3, day_2_uu.price * d, 3 * q * q * (1 - q))
day_3_udd = Node(8, 3, day_2_ud.price * d, 3 * q * (1 - q) * (1 - q))
day_3_ddd = Node(9, 3, day_2_dd.price * d, (1 - q) * (1 - q) * (1 - q))
day_4_uuuu = Node(10, 4, day_3_uuu.price * u, q * q * q * q)
day_4_uuud = Node(11, 4, day_3_uuu.price * d, 4 * q * q * q * (1 - q))
day_4_uudd = Node(12, 4, day_3_uud.price * d,
6 * q * q * (1 - q) * (1 - q))
day_4_uddd = Node(13, 4, day_3_udd.price * d,
4 * q * (1 - q) * (1 - q) * (1 - q))
day_4_dddd = Node(14, 4, day_3_ddd.price * d,
(1 - q) * (1 - q) * (1 - q) * (1 - q))
nodes = tree(day_0, day_1_u, day_1_d, day_2_uu, day_2_ud, day_2_dd,
day_3_uuu, day_3_uud, day_3_udd, day_3_ddd, day_4_uuuu,
day_4_uuud, day_4_uudd, day_4_uddd, day_4_dddd)
return nodes
def error(q, entry):
# Predicted nodes
tree_object = n.gen_tree(q, entry)
tree = [
tree_object.n0, tree_object.n1, tree_object.n2, tree_object.n3,
tree_object.n4, tree_object.n5, tree_object.n6, tree_object.n7,
tree_object.n8, tree_object.n9, tree_object.n10, tree_object.n11,
tree_object.n12, tree_object.n13, tree_object.n14
]
# Real-life data
real_life_data_day = [0, 1, 2, 3, 4]
real_life_data_price = [(fetch.close_data(entry)[i - 5]) for i in range(5)]
# Day 1 error
day_1 = tree[1:3]
error_day_1 = min(
[abs(real_life_data_price[1] - day_1[i].price)
for i in range(2)]) / real_life_data_price[1]
#print(error_day_1)
# Day 2 error
day_2 = tree[3:6]
error_day_2 = min(
[abs(real_life_data_price[2] - day_2[i].price)
for i in range(3)]) / real_life_data_price[2]
#print(error_day_2)
# Day 3 error
day_3 = tree[6:10]
error_day_3 = min(
[abs(real_life_data_price[3] - day_3[i].price)
for i in range(4)]) / real_life_data_price[3]
#print(error_day_3)
# Day 4 error
day_4 = tree[10:15]
error_day_4 = min(
[abs(real_life_data_price[4] - day_4[i].price)
for i in range(5)]) / real_life_data_price[4]
#print(error_day_4)
# Error set
errors = [error_day_1, error_day_2, error_day_3, error_day_4]
total_error = sum(errors)
#print(total_error)
return total_error
import fetch
import pandas as pd
renewable = ["run", "fslr", "cwen", "nee", "tsla", "plug"]
fossil = ["xom", "cvx", "btu", "arch", "cop"]
for stock in renewable:
fetch.close_data(stock).to_csv(stock + '.csv')
for stock in fossil:
fetch.close_data(stock).to_csv(stock + '.csv')
import fetch
close_data = fetch.close_data()
ratios = []
for index in range(len(close_data)-1):
ratios.append(close_data[index+1]/close_data[index])
def stock_price_ratios():
return ratios
def stock_price_ratios(entry):
ratios = []
close_data = fetch.close_data(entry)
for index in range(len(close_data) - 1):
ratios.append(close_data[index + 1] / close_data[index])
return ratios
import matplotlib.pyplot as plt import numpy as np import u_and_d_Hull_White as u_and_d import fetch from Hull_White_class import Node from datetime import datetime, timedelta import matplotlib.patches as mpatches u = u_and_d.u() d = u_and_d.d() day_0 = Node(0, 0, fetch.close_data()[-5], 1) day_1_u = Node(1, 1, day_0.price * u, 0.5) day_1_d = Node(2, 1, day_0.price * d, 0.5) day_2_uu = Node(3, 2, day_1_u.price * u, 0.25) day_2_ud = Node(4, 2, day_1_u.price * d, 0.5) day_2_dd = Node(5, 2, day_1_d.price * d, 0.25) day_3_uuu = Node(6, 3, day_2_uu.price * u, 0.125) day_3_uud = Node(7, 3, day_2_uu.price * d, 0.375) day_3_udd = Node(8, 3, day_2_ud.price * d, 0.375) day_3_ddd = Node(9, 3, day_2_dd.price * d, 0.125) day_4_uuuu = Node(10, 4, day_3_uuu.price * u, 0.0625) day_4_uuud = Node(11, 4, day_3_uuu.price * d, 0.25) day_4_uudd = Node(12, 4, day_3_uud.price * d, 0.375) day_4_uddd = Node(13, 4, day_3_udd.price * d, 0.25) day_4_dddd = Node(14, 4, day_3_ddd.price * d, 0.0625) tree = [day_0, day_1_u, day_1_d, day_2_uu, day_2_ud, day_2_dd, day_3_uuu, day_3_uud, day_3_udd, day_3_ddd, day_4_uuuu, day_4_uuud, day_4_uudd, day_4_uddd, day_4_dddd] fig,ax = plt.subplots()
import numpy as np import u_and_d_General as u_and_d import fetch from Hull_White_class import Node u = u_and_d.u() d = u_and_d.d() day_0 = Node(0, 0, fetch.close_data()[-5], 1) day_1_u = Node(1, 1, day_0.price * u, u) day_1_d = Node(2, 1, day_0.price * d, d) day_2_uu = Node(3, 2, day_1_u.price * u, u * u) day_2_ud = Node(4, 2, day_1_u.price * d, u * d) day_2_dd = Node(5, 2, day_1_d.price * d, d * d) day_3_uuu = Node(6, 3, day_2_uu.price * u, u * u * u) day_3_uud = Node(7, 3, day_2_uu.price * d, u * u * d) day_3_udd = Node(8, 3, day_2_ud.price * d, u * d * d) day_3_ddd = Node(9, 3, day_2_dd.price * d, d * d * d) day_4_uuuu = Node(10, 4, day_3_uuu.price * u, u * u * u * u) day_4_uuud = Node(11, 4, day_3_uuu.price * d, u * u * u * d) day_4_uudd = Node(12, 4, day_3_uud.price * d, u * u * d * d) day_4_uddd = Node(13, 4, day_3_udd.price * d, u * d * d * d) day_4_dddd = Node(14, 4, day_3_ddd.price * d, d * d * d * d)
def plot(q, entry):
u = u_and_d.u(q, entry)
d = u_and_d.d(q, entry)
tree_object = n.gen_tree(q, entry)
tree = [
tree_object.n0,
tree_object.n1,
tree_object.n2,
tree_object.n3,
tree_object.n4,
tree_object.n5,
tree_object.n6,
tree_object.n7,
tree_object.n8,
tree_object.n9,
tree_object.n10,
tree_object.n11,
tree_object.n12,
tree_object.n13,
tree_object.n14
]
fig,ax = plt.subplots()
day = [i.day for i in tree]
price = [i.price for i in tree]
#scale = [1000*abs(i.probability) for i in tree]
scale = 150
#colors = [i.probability for i in tree]
colors = 'royalblue'
colors_2 = 'grey'
colors_3 = 'white'
colors_4 = 'black'
colors_5 = 'red'
marker = '.'
font = 'Tahoma'
# Add arrows
for i in tree[0:10]:
x_tail = i.day
y_tail = i.price
x_head = i.day + 1
y_head_u = i.price * u
y_head_d = i.price * d
dx = x_head - x_tail
#dy_u = y_head_u - y_tail
#dy_d = y_head_d - y_tail
arrow_u = mpatches.FancyArrowPatch((x_tail, y_tail),(x_head, y_head_u), mutation_scale=15, zorder=1, color=colors_2)
arrow_d = mpatches.FancyArrowPatch((x_tail, y_tail),(x_head, y_head_d), mutation_scale=15, zorder=1, color=colors_2)
ax.add_patch(arrow_u)
ax.add_patch(arrow_d)
# Plot the chart
ax.scatter(day,price, s=scale, c=colors, marker=marker, zorder=2)
# Add rectangles
width = 0.5
height = (tree_object.n10.price-tree_object.n14.price)/15
for x,y in zip(day,price):
ax.add_patch(mpatches.Rectangle(xy=(x-width/2, y-height/2+(tree_object.n1.price-tree_object.n0.price)/2),
width=width,
height=height,
linewidth=1,
color=colors))
# Zip joins x and y coordinates in pairs
for x,y in zip(day,price):
label = "{:.2f}".format(y)
# Label each point with the price
ax.annotate(label, # this is the text
(x,y+(tree_object.n1.price-tree_object.n0.price)/2), # this is the point to label
va='center',
ha='center',
c=colors_3,
fontname=font) # horizontal alignment can be left, right or center
# Expected value plot
real_life_data_day = [0,1,2,3,4]
day_0_ev = tree[0].price
day_1_ev = (sum(tree[i].price*tree[i].probability for i in range(1,3)))
day_2_ev = (sum(tree[i].price*tree[i].probability for i in range(3,6)))
day_3_ev = (sum(tree[i].price*tree[i].probability for i in range(6,10)))
day_4_ev = (sum(tree[i].price*tree[i].probability for i in range(10,15)))
ev = [day_0_ev,day_1_ev,day_2_ev,day_3_ev,day_4_ev]
print(ev)
x = np.array(real_life_data_day)
y = np.array(ev)
m, b = np.polyfit(x, y, 1)
print(m)
ax.scatter(real_life_data_day,ev, s=scale, c=colors_5, marker=marker, zorder=2)
ax.plot(x, m*x + b, c=colors_5)
label2 = str(round(m,3)) + "(x)+ " + str(round(b,2))
print(label2)
ax.annotate(label2, # this is the text
(2,(tree_object.n14.price)), # this is the point to label
va='center',
ha='center',
c=colors_5,
fontname=font) # horizontal alignment can be left, right or center
# Real-life data
real_life_data_price = [(fetch.close_data(entry)[i-5]) for i in range(5)]
# Plot real-life data as a line plot
ax.plot(real_life_data_day,real_life_data_price, c=colors_4, zorder=4)
plt.xticks(np.arange(0,5,1))
stock = fetch.stock(entry)
start_day = datetime.today() - timedelta(days=4)
d1 = start_day.strftime("%B %d, %Y")
ax.set_xlabel('Days past ' + d1, fontname=font, weight='bold')
ax.set_ylabel('Price ($)', fontname=font, weight='bold')
ax.set_title(stock + ' Predicted Close Price Over Time', fontname=font, weight='bold')
fig.tight_layout()
plt.show()
import fetch
import pandas
fetch.close_data("plug").to_csv('stock.csv')
label = "{:.2f}".format(y)
# Label each point with the price
ax.annotate(label, # this is the text
(x,y+(n.day_1_u.price-n.day_0.price)/2), # this is the point to label
va='center',
ha='center',
c=colors_3,
fontname=font) # horizontal alignment can be left, right or center
# Real-life data
real_life_data_day = [0,1,2,3,4]
real_life_data_price = [(fetch.close_data()[i-5]) for i in range(5)]
# Plot real-life data as a line plot
ax.plot(real_life_data_day,real_life_data_price, c=colors_4, zorder=4)
plt.xticks(np.arange(0,5,1))
stock = fetch.stock()
start_day = datetime.today() - timedelta(days=4)
d1 = start_day.strftime("%B %d, %Y")
ax.set_xlabel('Days past ' + d1, fontname=font, weight='bold')
ax.set_ylabel('Price ($)', fontname=font, weight='bold')
ax.set_title(stock + ' Predicted Close Price Over Time', fontname=font, weight='bold')
fig.tight_layout()