Leverage Apache Spark During PyMC Sampling

[sldough21]

Is it possible to efficiently incorporate PyMC sampling within a Spark cluster? Curious to know if anyone has any experience with this.