def pdf_add():
global pdf_file, pdf_name_history, pdf_name
if pdf_name_history == pdf_name.get():
pdf_file.add()
pdf_file.save()
else:
pdf_name_history = pdf_name.get()
pdf_file = pdf.PDF("pdf/" + pdf_name.get())
pdf_file.add()
pdf_file.save()
import stocks
from statsmodels.tsa.arima_model import ARIMA
import matplotlib.pyplot as plt
import pdf
import pandas as pd
pr = pdf.PDF("pdf/arima_input.pdf")
def input_pdq():
x = tuple(
input(stocks.get_only_name() +
" : Enter, p d q : ").strip().split(" "))
if len(x) < 3:
x = (1, 1, 1)
y = [int(i) for i in x]
return y
def arima_man_forecast(ts, f, order=None):
confirm = False
if order is None:
confirm = True
train = ts[:int(len(ts) * f)]
test = ts[int(len(ts) * f):]
if order is None:
order = input_pdq()
model = ARIMA(train, order)
model_fit = model.fit(disp=-1)
plt.plot(test)
plt.plot(train)
import stocks
from pmdarima.arima import auto_arima
import matplotlib.pyplot as plt
import pdf
import pandas as pd
pr = pdf.PDF("pdf/auto_arima.pdf", True)
def arima_auto_forecast(ts, f):
train = ts[:int(len(ts) * f)]
test = ts[int(len(ts) * f):]
plt.plot(train)
plt.plot(test)
model = auto_arima(train)
prd = model.predict(n_periods=len(test))
plt.plot(prd)
print(prd)
plt.show()
while True:
arima_auto_forecast(stocks.stock["Close Price"], 0.6)
if not stocks.next_stock():
break
pr.save()
import pandas as pd
import matplotlib.pyplot as plt
import cvr
import stocks
import pdf
def frm(s):
s = str(s)
while len(s) < 8:
s += " "
return s
pr = pdf.PDF("pdf/savex.pdf")
coverWidth = 90
while True:
c = cvr.Cover(stocks.stock["Close Price"], coverWidth)
x = list()
y = list()
while c.move_forward():
# print(frm(c.min) + " " + frm(c.max) + " " + c.get_cover())
x.append((c.curr - c.min) / (c.max - c.min))
y.append(c.delta)
plt.title(stocks.get_name().split(".")[0].split("/")[1])
plt.xlabel("(curr-min)/(max-min) | max,min in past " + str(coverWidth) +
" days")
plt.ylabel("curr-price_on_previous_day")
plt.scatter(x, y, s=2)
pr.add()
plt.clf()
screen_width = win.winfo_screenwidth()
win.minsize(int(screen_width / 5), int(screen_height / 3))
lb = tk.Listbox(win)
files = os.listdir("stkdata/")
lb.configure(justify=tk.CENTER)
fill_list()
lb.bind("<Double-1>", list_click)
lb.pack()
# --------------------------------
pdf_frame = tk.Frame(win)
pdf_frame.pack()
pdf_name_history = "temp.pdf"
pdf_file = pdf.PDF("pdf/" + pdf_name_history)
pdf_name = tk.Entry(pdf_frame)
pdf_name.insert(0, pdf_name_history)
pdf_name.pack(side=tk.LEFT)
def pdf_add():
global pdf_file, pdf_name_history, pdf_name
if pdf_name_history == pdf_name.get():
pdf_file.add()
pdf_file.save()
else:
pdf_name_history = pdf_name.get()
pdf_file = pdf.PDF("pdf/" + pdf_name.get())
pdf_file.add()
pdf_file.save()
import stocks
import math
import numpy as np
import stock_iterator
import matplotlib
from matplotlib import pyplot as plt
from scipy.stats import norm
import pdf
import pandas as pd
import math
import mylib
matplotlib.rcParams["lines.linewidth"] = 0.5
train_data_percent = 0.9
# Load the data, and make training and testing sets.
pdf1 = pdf.PDF("pdf/MKV.pdf")
data = pd.Series(stocks.get_by_name("TCS"))
train_test_divider_index = int(train_data_percent * len(data))
train_data = data[:train_test_divider_index]
test_data = data[train_test_divider_index:]
train_length = len(train_data)
test_length = len(test_data)
plt.plot(train_data)
plt.plot(test_data)
# Show trend and scale on the graph
trend_polyfit = np.poly1d(np.polyfit(train_data.index, train_data, 3))
trend_series = pd.Series(trend_polyfit(data.index), index=data.index)
plt.plot(trend_series)
plt.title("TCS")
#!/usr/bin/python
import sys
sys.path += ["../scripts/"]
import pdf
# a test for transparency groups:
# form xobjects used for doing transparency groups can do savestate (q)
# without ever needing to do a corresponding restorestate (Q) because
# their content stream is self-contained.
#
# Test that this doesn't confuse the pdf reader.
file = pdf.PDF()
page = file.add_page(612, 100)
group1 = file.create_object("/XObject", "/Form")
group1.stream = """
0.0 1.0 0.0 rg
0.0 0.0 0.0 RG
10 10 m 70 10 l 70 70 l 10 70 l 10 10 l f
10 10 m 70 10 l 70 70 l 10 70 l 10 10 l s
0.0 0.0 1.0 rg
0.0 0.0 0.0 RG
30 30 m 90 30 l 90 90 l 30 90 l 30 30 l f
30 30 m 90 30 l 90 90 l 30 90 l 30 30 l s
1.0 0 0 1.0 1000 1000 cm q
1.0 0 0 1.0 1000 1000 cm q
1.0 0 0 1.0 1000 1000 cm q
1.0 0 0 1.0 1000 1000 cm q
"""
def __init__(self, name):
self.name = name
self.btnct = 0
self.curr_slab = None
self.pdf1 = pdf.PDF("pdf/transition_" + name + ".pdf")
import stocks
from statsmodels.tsa.arima_model import ARIMA
import matplotlib.pyplot as plt
import pdf
import pandas as pd
pr = pdf.PDF("pdf/arima_insample.pdf")
def input_pdq():
x = tuple(
input(stocks.get_only_name() +
" : Enter, p d q : ").strip().split(" "))
if len(x) < 3:
x = (1, 1, 1)
y = [int(i) for i in x]
return y
def arima_man_forecast(ts, f, order=None):
confirm = False
if order is None:
confirm = True
train = ts[:int(len(ts) * f)]
test = ts[int(len(ts) * f):]
if order is None:
order = input_pdq()
model = ARIMA(train, order)
model_fit = model.fit(disp=-1)
plt.plot(test)
plt.plot(train)
str(sg)[:5])
plt.hist(dif, bins=np.arange(-70, 70, 1), density=True)
plt.plot(np.arange(-70, 70, 1), norm.pdf(np.arange(-70, 70, 1), mu,
sg))
pd.add()
plt.clf()
return mu, sg
mu1, sg1 = ms(train[0:int(ntrain / 5)])
print(mu1, sg1)
mu2, sg2 = ms(train[int(ntrain / 5):2 * int(ntrain / 5)])
print(mu2, sg2)
mu3, sg3 = ms(train[2 * int(ntrain / 5):3 * int(ntrain / 5)])
print(mu3, sg3)
mu4, sg4 = ms(train[3 * int(ntrain / 5):4 * int(ntrain / 5)])
print(mu4, sg4)
mu5, sg5 = ms(train[4 * int(ntrain / 5):])
print(mu5, sg5)
mu = (mu1 + mu2 + mu3 + mu4 + mu5) / 5
sg = (sg1 + sg2 + sg3 + sg4 + sg5) / 5
brt = (mu + sg * sg) / 2
drt = (sg * sg - mu) / 2
print(brt, drt)
pd = pdf.PDF("pdf/ab.pdf")
# f(stocks.get_by_name("TCS"), "TCS")
stock_iterator.iterate(f)
pd.save()