CLMclement/dlsa

Distributed Least Squares Approximations

dlsa

Distributed Least Squares Approximation implemented with Apache Spark

System Requirements

• Spark >= 2.3.1

• Python >= 3.7.0

  • scikit-learn >= 0.21.2
  • rpy2 >= 3.0.4

• R >= 3.5

  • lars

  See setup.py for detailed requirements.

Run the PySpark code on the Spark platform

  ./bash/spark_dlsa_run.sh

or simply run

  ./logistic_dlsa.py

References

• Zhu, X., Li, F., & Wang, H., (2019) Least Squares Approximation for a Distributed System. Working Paper.