dipanjannag

dipanjannag

import java.io.InputStream

import org.apache.spark.sql.execution.datasources.jdbc.JdbcUtils
import org.apache.spark.sql.{ DataFrame, Row }

import org.postgresql.copy.CopyManager
import org.postgresql.core.BaseConnection

val jdbcUrl = s"jdbc:postgresql://..." // db credentials elided
val connectionProperties = {

dipanjannag

var AWS = require('aws-sdk');
exports.handler = (event, context, callback) => {
    AWS.config.region = 'us-east-1';
    var lexruntime = new AWS.LexRuntime();

    var params = {
        botAlias: "prod",
        botName: "OrderFlowers",
        inputText: event["body-json"].message,
        userId: event["body-json"].uuid,

dipanjannag

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": [
                "logs:CreateLogGroup",
                "logs:CreateLogStream",
                "logs:PutLogEvents"
            ],

dipanjannag

def trade_points(data,max_spread, pair=(136308228,2714625)):
    test_bse = data[data.instrument_token==pair[0]]
    test_nse = data[data.instrument_token==pair[1]]
    #print test_bse.head()
    #print test_nse.head()
    test_nse['market'] = 'nse'
    test_bse['market'] = 'bse'
    bse_bid = test_bse.iloc[0]['depth_buy_price_0 ']
    nse_bid = test_nse.iloc[0]['depth_buy_price_0 ']
    bse_offer = test_bse.iloc[0]['depth_sell_price_0 ']

dipanjannag

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [

dipanjannag

#Return on quantity traded, across different sell value
x_b = range(10, 250, 10)
y_i = []
for i in range(10):
    y_i.append([(0.2*i*x)-2*charge(1001,1001+(0.2*i),x) for x in x_b])
    
for i in range(10):
    plt.plot(x_b, y_i[i], '-', label = 'Sell @'+str((0.2*i)+1001))

plt.legend()

dipanjannag

import matplotlib.pyplot as plt
from pylab import rcParams
#Lets plot quantity against charge of arbitrage
rcParams['figure.figsize'] = 15, 10
qty = range(10, 2500, 10)
# multiply by 2. Because, to do arbitrage in intraday we need to do 4 trades
# i.e. 2 complete trade cycle
c_y = [2*charge(1000,1000.5, x) for x in qty]
fig, ax = plt.subplots()
ax.plot(qty,c_y,'-')

dipanjannag

## Function to calculate charges paid on a full cycle trade
## Asumptions:
## 1. Transection charge: High value transaction happened on NSE
##    i.e. the sell side in normal arbitrage. So for large quantity
##    it will be higher. So we might calculate higher brokerage, but 
##    never lower. Better to have a safety margin.
def charge(buy, sell, qty):
    buy = buy*qty
    sell = sell*qty
    turnover = buy+sell

dipanjannag

def arbitrage_opportunity(nse_id, bse_id):
    import pandas as pd
    import datetime
    tick_data = pd.read_csv('log_raw.csv')
    tick_data.timestamp = pd.to_datetime(tick_data.timestamp)
    # filter-out pre-market data. 03:45 UTC means 9:15 IST. 9:00 to 9:15 is pe-market session
    tick_data = tick_data.loc[tick_data.timestamp > '2017-06-19 03:45:00.000']
    stock_nse = tick_data[tick_data.instrument_token == nse_id]
    stock_bse = tick_data[tick_data.instrument_token == bse_id]
    print "bse feed mean interval",stock_bse.timestamp.diff().fillna(0).mean()