Update File Reader. Add limit instance as stopping condition

[nhnminh]

This pull request has 1 change in File Reader:

Change:

• Add a parameter for limit number of instances -i in FileReader.
• Stop Spark Context if the number of instances reaches the limit, by using ssc.stop(stopSparkContext = false, stopGracefully = false)

Reason: Stopping Condition, used for benchmarking (computing running time).

Test:

We could run the code as follow:

./spark.sh "200 EvaluatePrequential  -l (trees.HoeffdingTree -l 0 -t 0.05 -g 200 -o) -s (FileReader -f ../data/electNormNew.arff -k 4000 -d 10 -i 30000)"  1> result.res 2> log.log

Here, we limit the number of instances to be 30000