Emotion Categorization
Running : We are using sbt. If you have trouble with the dependencies or any general questions, contact Ausitn
> sbt
> compile
> project
> run
As an example
> sbt
> compile
> project polarity
> run-main main.scala.PolarityDistribution Tweet-Data/Tiny-Labeled.csv Tweet-Data/Tiny-Polarity.ser
> run-main main.scala.CalculatePolarities -s Tweet-Data/Tiny-Polarity.ser Tweet-Data/Tiny-Unlabeled.csv Tweet-Data/Tiny-features.txt
If you are running a very large dataset, you should modify your sbt configuration to increase RAM, ie,
#!/bin/sh
test -f ~/.sbtconfig && . ~/.sbtconfig
exec java -Xmx400G -Xms20G -XX:MaxPermSize=40G ${SBT_OPTS} -jar ~/bin/sbt-launch.jar "$@"
Todo : describe how to run liblinear / libsvm
Testing :
We are using scalatest for testing. From the root directory :
> sbt test
Should run and compile all test. You can also test within a specific project, or use test-only to call a specific test.
Todo :
- Unit Tests
- Streaming Support
- Documentation Auto-generation
- Maven for external jar/lib handling