Logistic Regression with R and Hadoop

Question

We are using rmr and rhadoop package of RevoR . Can we perform linear regression on an entire data set in hadoop without the need to implement the linear regression algorithm in map reduce

or

Is rmr and rhadoop, just used for summarizing the data and then running regression on the summarized data locally?

Answer 1

Neither. You can implement any sort of algorithm in rmr, not only summaries. However, it's not an algorithm package, as such it doesn't include linear regression or clustering or anything of that sort, much like packages parallel or snow.