def Update_trends(self):
# Change to data directory
os.chdir(self.home_dir + '\\Data')
# define date
my_cryptory = Cryptory(from_date="2018-05-01")
# pull data from google trends
df = my_cryptory.get_google_trends(kw_list=["cryptokitties"])
# save as csv
df.to_csv('trends.csv', index=0)
print("\nUpdated google trends data, saved in trends.csv\n")
# change directory back to original
os.chdir(self.home_dir)
def Update_ETH_price(self):
# Change to data directory
os.chdir(self.home_dir + '\\Data')
# define date to pull from
my_cryptory = Cryptory(from_date="2018-05-01")
# get historical ethereum prices from coinmarketcap
df = my_cryptory.extract_coinmarketcap("ethereum")
# save as csv
df.to_csv('ETH_price.csv', index=0)
print("\nUpdated Ethereum price data, saved in ETH_price.csv\n")
# change directory back to original
os.chdir(self.home_dir)
def __init__(self,
date_from,
date_to,
days=None,
bitinfo_list=None,
google_list=None,
prediction_ticker='ltc'):
if bitinfo_list is None:
bitinfo_list = ['btc', 'eth']
if google_list is None:
google_list = ['Litecoin']
cryp_obj = Cryptory(from_date=date_from, to_date=date_to)
fin_table = cryp_obj.extract_bitinfocharts(bitinfo_list[0])
if len(bitinfo_list) > 1:
for i in range(1, len(bitinfo_list)):
temp_table = cryp_obj.extract_bitinfocharts(bitinfo_list[i])
temp_table = temp_table.drop(columns='date')
fin_table = pd.concat([fin_table, temp_table],
axis=1,
join_axes=[temp_table.index])
prediction_table = cryp_obj.extract_bitinfocharts(prediction_ticker)
prediction_table = prediction_table.drop(columns='date')
if google_list:
google_table = cryp_obj.get_google_trends(google_list)
self.final_table = pd.concat(
[fin_table, google_table, prediction_table],
axis=1,
join_axes=[prediction_table.index])
else:
self.final_table = pd.concat([fin_table, prediction_table],
axis=1,
join_axes=[prediction_table.index])
if days is not None:
self.days_out = days
X_test = []
Y_test = []
def get_mean():
global Y_mean
return Y_mean
def get_std():
global Y_std
return Y_std
#####GET GOOGLE TRENDS#####
my_cryptory = Cryptory(from_date="2015-08-25", to_date="2018-04-04")
trend1 = my_cryptory.get_google_trends(kw_list=['bitcoin'])
trend2 = my_cryptory.get_google_trends(kw_list=['ethereum'])
def generatedata_v1(isTrain, batch_size):
"""
Return data regarding cyrptocurrencies.
"""
# window for encoder and decoder's predictions.
seq_length = 40
global Y_train
global X_train
global X_test
import re
import plotly.graph_objects as go
import time
# initialise object
# pull data from start of 2017 to present day
from_date = "2013-04-28"
to_date = "2020-03-21"
to_date_bitcoin = "2020-03-20"
from_date_cmp = time.strptime(from_date, "%Y-%m-%d")
to_date_cmp = time.strptime(to_date, "%Y-%m-%d")
lowest_date = ''
highest_date = ''
my_cryptory = Cryptory(from_date=from_date, to_date=to_date, timeout=100)
coin_col = False
coin = "bitcoin"
def extract_coinmarketcap(coin, coin_col=False):
try:
output = pd.read_html(
"https://coinmarketcap.com/currencies/{}/historical-data/?start={}&end={}"
.format(coin, from_date.replace("-", ""),
to_date_bitcoin.replace("-", "")))[2]
except:
# future versions may split out the different exceptions (e.g. timeout)
raise
Created on 4 mrt. 2020
@author: stan
'''
# load package
from cryptory import Cryptory
import datetime
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import re
import plotly.graph_objects as go
# initialise object
# pull data from start of 2017 to present day
my_cryptory = Cryptory(from_date="2015-12-31", to_date="2020-03-21")
dates = list(reversed(my_cryptory.extract_bitinfocharts("btc")["date"]))
#print(dates)
priceBTC = list(reversed(
my_cryptory.extract_bitinfocharts("btc")["btc_price"]))
#print(my_cryptory.get_google_trends(kw_list=['bitcoin']))
GoogleTrends = list(
reversed(my_cryptory.get_google_trends(kw_list=['bitcoin'])["bitcoin"]))
#Subreddit
RedditData = my_cryptory.extract_reddit_metrics(
"bitcoin", "subscriber-growth-perc")["subscriber_growth_perc"]
#print("Length dates: {} ".format(len(dates)))
#print("Length priceBTC: {} ".format(len(priceBTC)))
# load package
from cryptory import Cryptory
# initialise object
# pull data from start of 2017 to present day
my_cryptory = Cryptory(from_date="2017-01-01")
######## Basic Usage
# get historical bitcoin prices from coinmarketcap
my_cryptory.extract_coinmarketcap("bitcoin")
# get daily subscriber numbers to the bitcoin reddit page
my_cryptory.extract_reddit_metrics(subreddit="bitcoin",
metric="total-subscribers")
# google trends- bitcoin search results
my_cryptory.get_google_trends(kw_list=["bitcoin"])
# dow jones price (market code from yahoo finance)
my_cryptory.get_stock_prices(market="%5EDJI")
# USD/EUR exchange rate
my_cryptory.get_exchange_rates(from_currency="USD", to_currency="EUR")
# get historical commodity prices
my_cryptory.get_metal_prices()
######## Advanced Usage
# generate price correlation matrix
print('Ended portfolio in EUR: {}'.format(
locale.currency(EUROS, grouping=True)))
print('% Profit/Loss: {}%'.format(
round(((EUROS - startingPortfolio) / startingPortfolio) * 100), 2))
#Plotting results
y_axis_breakeven = []
for i in range(0, len(y_axis)):
y_axis_breakeven.append(startingPortfolio)
#date_1 = datetime.strptime(x_axis[0], "%Y-%m-%d")
#modifiedDate = date_1 + timedelta(days=1)
modifiedDate = x_axis[0]
my_cryptory = Cryptory(
from_date=modifiedDate,
to_date=presentDate) # First element of x_axis = start date of data
priceBTC = list(reversed(
my_cryptory.extract_bitinfocharts("btc")["btc_price"]))
print(my_cryptory.extract_bitinfocharts("btc"))
print("Length Portfolio: {} ".format(len(y_axis)))
print("Length BTC Pric: {} ".format(len(priceBTC)))
print("Length Break-even: {} ".format(len(y_axis_breakeven)))
print("Start date: {} ".format(x_axis[0]))
print("Start date modified +1: {} ".format(modifiedDate))
#Create dataframe
df = pd.DataFrame({'Portfolio': y_axis})
df['BTC Price'] = priceBTC
df['Break-even'] = y_axis_breakeven
from cryptory import Cryptory
import pandas as pd
my_cryptory = Cryptory(from_date="2017-01-01")
pd.options.display.max_rows = 6
all_coins_df = my_cryptory.extract_bitinfocharts("btc")
bit_info_coins = ['btc', 'eth', 'xrp', 'bch', 'ltc']
for coin in bit_info_coins[1:]:
all_coins_df = all_coins_df.merge(my_cryptory.extract_bitinfocharts(coin),
on='date',
bow='left')
import seaborn as sns
import matplotlib.pyplot as plt
import datetime
fig, ax1 = plt.subplots(1, 1, figsize=(10, 4))
ca = plt.get_cmap('gist_rainbow')
print(my_cryptory.head())
# Help on class Cryptory in module cryptory.cryptory:
my_cryptory.extract_poloniex(coin1="btc", coin2="eth")
from cryptory import Cryptory
import pandas as pd
from dtaidistance import dtw
import matplotlib.pyplot as plt
from dtaidistance import dtw_visualisation as dtwvis
from dtw_btc.templates import HeadAndShouldersTop, HeadAndShouldersBottom, TripleTopReversal, \
DoubleTopReversal, FallingWedge, DoubleBottomReversal, RisingWedge, TripleBottomReversal, \
BumpAndRunReversal, SymmetricalTriangle
from dtw_btc.utils import normalize, denormalize, normalize_function_paper
cryptory_from_date = Cryptory(from_date="2020-05-01")
history_price_btc = cryptory_from_date.extract_coinmarketcap(
"bitcoin").sort_values(by=['date'])
list_of_dates = list(history_price_btc['date'])
list_of_closes = list(history_price_btc['close'])
len_list_of_closes = len(list_of_closes)
patterns = [
HeadAndShouldersTop(),
HeadAndShouldersBottom(),
TripleTopReversal(),
DoubleTopReversal(),
DoubleBottomReversal(),
FallingWedge(),
RisingWedge(),
TripleBottomReversal(),
BumpAndRunReversal(),
SymmetricalTriangle()
from cryptory import Cryptory
import matplotlib.dates as mdates
import seaborn as sns
import matplotlib.pyplot as plt
import datetime
my_cryptory = Cryptory(from_date="2017-01-01")
# btc_google = my_cryptory.get_google_trends(kw_list=['bitcoin']).merge(
# my_cryptory.extract_coinmarketcap('bitcoin')[['date','close']], on='date', how='inner')
# eth_google = my_cryptory.get_google_trends(kw_list=['ethereum']).merge(
# my_cryptory.extract_coinmarketcap('ethereum')[['date','close']], on='date', how='inner')
ltc_google = my_cryptory.get_google_trends(kw_list=['litecoin']).merge(
my_cryptory.extract_coinmarketcap('litecoin')[['date', 'close']],
on='date',
how='inner')
iota_google = my_cryptory.get_google_trends(kw_list=['iota']).merge(
my_cryptory.extract_coinmarketcap('iota')[['date', 'close']],
on='date',
how='inner')
# ripple_google = my_cryptory.get_google_trends(kw_list=['ripple']).merge(
# my_cryptory.extract_coinmarketcap('ripple')[['date','close']], on='date', how='inner')
doge_google = my_cryptory.get_google_trends(kw_list=['Dogecoin']).merge(
my_cryptory.extract_coinmarketcap('Dogecoin')[['date', 'close']],
on='date',
how='inner')
class MyDataManager():
def __init__(self, date):
self.crypto_data = Cryptory(from_date = date)
self.bitcoin_data = self.crypto_data.extract_coinmarketcap("bitcoin")
self.sc = MinMaxScaler()
# Cleans the data to keep only the open column
def clean_data(self):
for col in self.bitcoin_data.columns:
if col != "open":
del self.bitcoin_data[col]
# Creates the train data set
def create_train_data(self):
self.clean_data()
training_set = self.bitcoin_data;
training_set = self.sc.fit_transform(training_set)
train_data = training_set[365:] # around 80% of the data is used for training
# Split the data into x and y
x_train, y_train = train_data[:len(train_data)-1], train_data[1:]
return self.format_to_3d(x_train), y_train
# Creates the test data set
def create_test_data(self):
test_set = self.bitcoin_data
test_set = self.sc.transform(test_set)
test_data = test_set[:364]
return test_data
# Plots the graph
def plot(self, real_value, model):
input = real_value
input = self.sc.fit_transform(input)
input = np.reshape(input, (364, 1, 1))
predicted_result = model.predict(input)
predicted_result = self.sc.inverse_transform(predicted_result)
print(real_value)
plt.plot(real_value, color='pink', label='Real Price')
plt.plot(predicted_result, color='blue', label='Predicted Price')
plt.title('Bitcoin Prediction')
plt.xlabel('Time')
plt.ylabel('Prices')
plt.legend()
plt.show()
# Format a 2D np.array to a 3D array
def format_to_3d(self, df_to_reshape):
reshaped_df = np.array(df_to_reshape)
return np.reshape(reshaped_df, (reshaped_df.shape[0], 1, reshaped_df.shape[1]))
def __init__(self, date):
self.crypto_data = Cryptory(from_date = date)
self.bitcoin_data = self.crypto_data.extract_coinmarketcap("bitcoin")
self.sc = MinMaxScaler()
# generate price correlation matrix
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
# load package
from cryptory import Cryptory
import datetime
# initialise object
# pull data from start of 2017 to present day
my_cryptory = Cryptory(from_date="2016-01-01")
# get historical bitcoin prices from coinmarketcap
my_cryptory.extract_coinmarketcap("bitcoin")
all_coins_df = my_cryptory.extract_bitinfocharts("btc")
# coins of interest
bitinfocoins = ["btc", "eth", "xrp", "bch", "ltc", "dash", "xmr", "doge"]
for coin in bitinfocoins[1:]:
all_coins_df = all_coins_df.merge(my_cryptory.extract_bitinfocharts(coin),
on="date",
how="left")
# date column not need for upcoming calculations
all_coins_df = all_coins_df.drop('date', axis=1)
corr = all_coins_df.pct_change().corr(method='pearson')
# fig, ax = plt.subplots(figsize=(7,5))
# sns.heatmap(corr,
# xticklabels=[col.replace("_price", "") for col in corr.columns.values],
# yticklabels=[col.replace("_price", "") for col in corr.columns.values],
# annot_kws={"size": 16})
# plt.show()