API Class Tutorial

The primary role of the API class is to request data from a Bitcoin exchange API, standardize the format of the data, and return the data as a pandas DataFrame. The class also provides additional features like ping_limit and auto_since which are examined in further detail below.

Note: You can see a complete script of this tutorial here and the complete function documentation here

##Add request jobs to API object

First you must create the API class and then add specific jobs to the class. Currently there are two types of jobs which the API can perform.

• ticker : This job will return a single row of ticker data which gives the last price, 24 hour high and low, the current bid and ask spread, and the 24 hour volume.

• trades: This job will return the most recent trade history including the price, volume, bid/ask, and time of the trade.

In order to add jobs you must use the add_job([exchange],[symbol],[type]) method. This method also allows you to set optional settings like limit, auto_since, and ping_limit.

>> import bitquant as bq
>> a = bq.api()
>> a.add_job('bitfinex','btcusd','ticker')
>> a.add_job('bitfinex','btcusd','trades')
>> a.add_job('btcchina','btccny','trades',limit=2)

Note that you can also use add_job method to override or remove any optional settings previous set

##Run API requests jobs

To actually request data from an Bitcoin exchange API, you must use the run([exchange],[symbol],[type]) method which will return a DataFrame. You do not need to include the optional settings when using the run method.

>> tick = a.run('bitfinex','btcusd','ticker')
>> print tick

      ask     bid    high    last    low             timestamp  \
0  231.27  231.26  231.66  231.26  226.1  1441397010.479910588   

           volume  exchange  symbol  
0  11866.70689331  bitfinex  btcusd  

>> trd = a.run('bitfinex','btcusd','trades')
>> print trd

        amount  exchange   price       tid   timestamp  type  symbol
0        0.275  bitfinex  231.26  11565643  1441397003  sell  btcusd
1        0.422  bitfinex  231.26  11565641  1441397003  sell  btcusd
2        0.422  bitfinex  231.26  11565639  1441397003  sell  btcusd
3        0.422  bitfinex  231.26  11565637  1441397003  sell  btcusd
4        0.422  bitfinex  231.26  11565635  1441397003  sell  btcusd
5        0.421  bitfinex  231.26  11565633  1441397003  sell  btcusd
6        0.422  bitfinex  231.26  11565631  1441397003  sell  btcusd
7        0.271  bitfinex  231.26  11565629  1441397003  sell  btcusd
8        0.151  bitfinex  231.26  11565627  1441396988  sell  btcusd

>> trd = a.run('btcchina','btccny','trades')
>> print trd

  amount   timestamp   timestamp_ms    price        tid  type exchange  symbol
0  0.037  1441397018  1441397018000  1473.66  679485368  sell   okcoin  btccny
1  0.012  1441397018  1441397018000  1473.62  679485370  sell   okcoin  btccny

##Optional Setting: limit

As demonstrated above, the limit setting will limit the number of rows returned by the API request. By default this option is turned off (limit='').

##Optional Setting: ping_limit

You may change the ping_limit for a particular job when you call the add_job method. The ping limit will enforce a provided number of seconds between API requests so your scripts to not overload the exchanges with requests. By default it is set to ping_limit=1.0 which is one request per second, but could be changed for example to ping_limit=0.5 which is two requests per second.

##Optional Setting: auto_since

The auto_since setting will automatically minimize the number of rows returned by the API request such that only new data is returned. Each exchange has slightly different support for this feature, some exchanges will limit rows based on the last trade ID and some will limit based on the last timestamp.

In the complete script for this tutorial, there is a demonstration of this feature which is displayed below.

a.add_job('okcoin','btccny','trades',auto_since='yes')
for i in range(10):
    print a.run('okcoin','btccny','trades')

Output sample:

[0 rows x 0 columns]
  amount   timestamp   timestamp_ms    price        tid  type exchange  symbol
0  0.089  1441397021  1441397021000  1473.56  679485451   buy   okcoin  btccny
1    0.4  1441397021  1441397021000  1473.58  679485453   buy   okcoin  btccny
2  0.048  1441397021  1441397021000  1473.58  679485455   buy   okcoin  btccny
3  0.256  1441397021  1441397021000  1473.58  679485457   buy   okcoin  btccny
4  0.437  1441397022  1441397022000  1473.58  679485470  sell   okcoin  btccny
5  0.264  1441397022  1441397022000  1473.58  679485473  sell   okcoin  btccny

[6 rows x 8 columns]
  amount   timestamp   timestamp_ms    price        tid type exchange  symbol
0    0.3  1441397024  1441397024000  1473.59  679485503  buy   okcoin  btccny
1  0.157  1441397024  1441397024000  1473.59  679485504  buy   okcoin  btccny

[2 rows x 8 columns]
  amount   timestamp   timestamp_ms    price        tid type exchange  symbol
0  0.042  1441397029  1441397029000  1473.58  679485560  buy   okcoin  btccny
1  0.258  1441397029  1441397029000  1473.58  679485562  buy   okcoin  btccny
2   0.06  1441397029  1441397029000  1473.58  679485563  buy   okcoin  btccny
3  0.295  1441397029  1441397029000  1473.58  679485565  buy   okcoin  btccny

[4 rows x 8 columns]
Empty DataFrame
Columns: []
Index: []