infoankitp

infoankitp

import com.google.common.hash.HashFunction
import com.google.common.hash.Hashing

val SEED = 123456;
val LOG2M = 15
val REGWIDTH = 5 
val EXPTHRESH = -1
val SPARSEON = true

// A User Defined Function which is used to initiate HLL for a String

infoankitp

final static int SEED = 123456;
final static int LOG2M = 15; // Number of Buckets  = 2^LOG2M
final static int REGWIDTH = 5;
final static int EXPTHRESH = -1;
final static boolean SPARSEON = true;
 
 
/**
 * Generates a HLL Object with numbers from 0 to n the parameter passed
 * @param n : Number of Elements in the HLL.

infoankitp

CREATE TABLE hll_test (page TEXT, user_id_hll bytea, session_id_hll bytea);

infoankitp

CREATE TABLE test_table (page TEXT, user_id TEXT, session_id TEXT);
INSERT INTO test_table VALUES ('A', '1', 'A'), ('B', '1', 'B') ,('A', '2', 'C');
/* Since we have already added the hll Extension,
 * We can now try running the following query to get the unique_user_count and unique_session_count */
SELECT hll_cardinality(hll_add_agg(hll_hash_text(user_id))) as unique_user_count, hll_cardinality(hll_add_agg(hll_hash_text(session_id))) as unique_session_count FROM test_table;

infoankitp

CREATE EXTENSION hll;

infoankitp

git clone https://github.com/citusdata/postgresql-hll.git
cd postgresql-hll
sed -i -e 's,lstdc++,L/usr/lib,g' Makefile # This is done to include LD libraries, Also please make sure gcc is downloaded on your system.
PG_CONFIG=/Library/PostgreSQL/9.6/bin/pg_config make
sudo make install

infoankitp

Page	UserId	SessionId
index.html	1	A
1.html	1	B
index.html	2	C

Count(Distinct userId)	Count(Distinct SessionId)
2	3

infoankitp

import time
import boto3
import urllib
import datetime
import os
import json
import requests
import traceback
from PIL import Image
from io import BytesIO

infoankitp

import json
def lambda_handler(event, context):
    print( "Event : " , event)  
    #This will print the event passed to lambda_handler in     
    #Cloudwatch Logs
    print( "Context : " , context) 
    #This will print the context passed to lambda_handler in  
    #Cloudwatch #Logs
    name = ""
    if 'queryStringParameters' in event: