Project report included. Name of the file: Project_Report_NYTimes_Analysis.pdf

#Installation Instructions

All the installations are to be carried out as root user

1. Install Hadoop on all the nodes as specified in the Instructions document provided in the class.
Set all the masters and slaves.
**Note: The Java version has to be 8. Otherwise, this project will not run**
2. Increase the JVM heap space to 4GB. In /usr/lib/hadoop/libexec/hadoop-config.sh, 
set JAVA_HEAP_MAX=-Xmx4096m. The 'JAVA_HEAP_MAX' variable has to be updated in 2 places in the file.
This has to be done on all nodes.
3. Install MongoDB on the master node using the commands given below:
   - sudo apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv 7F0CEB10
   - echo 'deb http://downloads-distro.mongodb.org/repo/ubuntu-upstart dist 10gen' | sudo tee /etc/apt/sources.list.d/mongodb.list
   - sudo apt-get update
   - sudo apt-get install -y mongodb-org
4. In /etc/mongod.conf, change 'bind_ip = 127.0.0.1' to '#bind_ip = 127.0.0.1'.
   Basically, comment it out so that mongo listens on all IP addresses.
5. Restart MongoDB. -> $ service mongod restart
6. Install mongoengine as show below:
   - apt-get install python-pip
   - pip install mongoengine
7. Install R as shown below:
   - sudo apt-get update
   - sudo apt-get install r-base
   - Install g++ if not already installed.
   - Login to R. ($ R) and execute the following 2 commands:
	install.packages("ggplot2")
	install.packages("reshape2")
   - Exit R by typing 'q()'.

#Execution Instructions

sudo su - root
cd /home/ubuntu

Copy/move all the above directories and files to /home/ubuntu

- Download StanfordCoreNLP using:
  sudo wget https://www.dropbox.com/s/mvbct0xvp5hptef/stanford-corenlp-3.5.0-models.jar?dl=1
  sudo mv stanford-corenlp-3.5.0-models.jar\?dl\=1 stanford-corenlp-3.5.0-models.jar
  sudo mv stanford-corenlp-3.5.0-models.jar /home/ubuntu/lib

- Copy 'mongo-java-driver-2.12.4.jar' and 'mongo-hadoop-core-1.3.0.jar' from '/home/ubuntu/lib'
directory to '/usr/lib/hadoop/lib'. This is to be done on all nodes.
- Restart hadoop services on all nodes.

cd CompilerCode
chmod 755 compiler.sh
./compiler.sh
cd /home/ubuntu

chmod 755 DataLoader.sh
chmod 755 NYTimesSentimentAnalysis.sh
chmod 755 StockSentimentCorrelation.sh

./DataLoader.sh
./NYTimesSentimentAnalysis.sh <IPAdress of the server running MongoDB instance> <Country>
./StockSentimentCorrelation.sh <IPAdress of the server running MongoDB instance> <Company>

- Country and Company names should be specified in CamelCase. (First letter uppercase)
- Countries available: China, Japan, Germany.
(Due to space constraints, we have uploaded the data only for these 3 countries so
that it can be used to test the code.)
- Graphs for China, Germany, India, Israel, Japan, Russia, Iran and Microsoft are available in
the 'Graphs' directory. They can be used for verification.
- Companies available: Microsoft, Amazon

The output graph is a PDF stored in /home/ubuntu (format: country_sentiment.pdf or company_combined.pdf)