environment

- MacOS Sierra 10.12.6
- Python 3.5.3 :: Anaconda 4.4.0 (x86_64)
- Jupyter 4.3.0

background

――Zaif exchange is a valuable virtual currency exchange with virtual currencies that can be bought and sold in pairs with Japanese yen only here, such as NEM and Monacoin, as well as Bitcoin. --The Zaif exchange can make its website so heavy that it can't even log in, let alone trade, if the volume of transactions surges, such as when Bitcoin-related (mainly negative) news is reported.

problem

--If you cannot log in to the Zaif exchange, it will be accurate (especially) for cryptocurrencies that are not traded in pairs with Japanese yen other than the Zaif exchange, or are traded only on the Zaif exchange such as Zaif tokens. You will not be able to know the price movement.

Solution

--The Zaif exchange publishes an API that allows you to obtain trading price information and make transactions, and even if the website is heavy, the API often survives. Therefore, if you periodically acquire the trading price with the API and display the chart, you can check the price movement even if you can not log in to the website. -** Click here for the created one (chart is updated every 5 minutes): ** https://zaifcharts.hateblo.jp/ --Click here for how to buy and sell with API after checking the chart: Buy and sell virtual currency using Zaif API

specification

Create a script that acquires trading price information from the Zaif exchange and writes it to a csv file, and a script that creates a chart from the written csv file, and execute them regularly with cron.

Obtain trading price information from Zaif exchange

--Import required modules such as pandas, numpy, json, etc.

import pandas as pd
from pandas import Series,DataFrame
import numpy as np
from datetime import datetime
import time
import json
import requests
import csv
import sys, codecs

--Get the current time (to make it the horizontal axis of the chart)

date = datetime.now().strftime("%Y/%m/%d %H:%M:%S")

--Get the latest trading price using the API of Zaif exchange (only BTC is described below, the same applies to other virtual currencies)

response = requests.get('https://api.zaif.jp/api/1/last_price/btc_jpy')
if response.status_code != 200:
    raise Exception('return status code is {}'.format(response.status_code))
res_dict = json.loads(response.text)
btc_price = res_dict['last_price']
#[Current time, selling price]Create a list of
btc_list = [date, btc_price]

--Write the acquired trading price to a csv file (only BTC is described below, the same applies to other virtual currencies)

f = open("/var/www/html/zaif/btc.csv", "a")
writer = csv.writer(f, lineterminator='\n')
writer.writerow(btc_list)
f.close()

If you execute a script that includes an operation to write to a file with cron, the following error may be output and you may not be able to execute it.

UnicodeDecodeError: 'ascii' codec can't decode byte 0xe3 in position 720: ordinal not in range(128)

This time, I avoided the error by writing ʻexport LANG = ja_JP.UTF-8;` in crontab as shown below.

0,5,10,15,20,25,30,35,40,45,50,55 * * * * export LANG=ja_JP.UTF-8; runipy /home/user/Get_Zaif_data.ipynb

Create a chart from the acquired trading price information

--Import required modules such as matplotlib

import pandas as pd
from pandas import Series,DataFrame
import numpy as np
from datetime import datetime
import time
import json
import requests
import csv
import matplotlib.pyplot as plt
from matplotlib.pylab import rcParams
%matplotlib inline
import seaborn as sns
sns.set_style('whitegrid')
rcParams['figure.figsize'] = 20, 10
from matplotlib.dates import DateFormatter

--Read csv file and create Dataframe (The following describes only BTC, the same applies to other virtual currencies)

btc_df = pd.read_csv("/var/www/html/zaif/btc.csv", names = ("date", "BTC price"))

--Extract only the period for creating the chart (3 days in the example below)

btc_df = btc_df[-864:-1]

--Mold the date column of btc_df into hh: mm format

date_df = btc_df["date"].str[10:16]
btc_df = pd.concat([date_df, btc_df["BTC price"]], axis=1)

--Calculate BTC bollinger band (The following describes only BTC, the same applies to other virtual currencies)

rm = btc_df.set_index("date").rolling(window=24).mean()
rstd = btc_df.set_index("date").rolling(window=24).std()
upper_band = rm + rstd * 2
lower_band = rm - rstd * 2

For more information, see this page: http://www.madopro.net/entry/MachineLearningForTrading *

--Create a chart from Dataframe (The following describes only BTC, the same applies to other virtual currencies)

ax = btc_df.set_index("date").plot()
rm.columns = ["Rolling mean"]
rm.plot(ax=ax, color="#8FBC8F")
upper_band.columns = ["Upper band"]
upper_band.plot(ax=ax, color="#F4A460")
lower_band.columns = ["Lower band"]
lower_band.plot(ax=ax, color="#F4A460")
plt.legend(loc='best',
           fontsize=14,
           borderaxespad=0.,)
plt.tick_params(labelsize=14)
plt.savefig("/var/www/html/zaif/btc.png ")

--Created chart

Summary

--By creating a chart from the trading price information obtained from the API, you can now check the price movements of virtual currencies traded on the Zaif exchange even if you cannot log in to the Zaif exchange website. --Even if you can log in to the Zaif exchange website, it is convenient to have a page that can list the charts of each virtual currency (on the Zaif exchange website, you can see the charts of multiple virtual currencies at once. Can't).

Get the trading price of virtual currency and create a chart with API of Zaif exchange