November 2017
Beginner to intermediate
366 pages
7h 59m
English
We are not concerned about the batch layer for this exercise. The batch layer is responsible for keeping the master data intact. Since it has a very low latency, we have the speed layer to process the continuous stream of data.
Now that we have a good idea of the overall architecture of our system, let us proceed to develop the stream clustering algorithm. For the cluster algorithm development purpose, we are going to leverage the DSD simulators in the stream package to act as our sensors in the digital control system.
DSD data generation:
> sensor.stream <- DSD_Gaussians(k = 2, d = 4, mu = rbind(c(200,77,20,750),c(196,80,16,790)))> sensor.streamMixture of GaussiansClass: DSD_Gaussians, DSD_R, DSD_data.frame, DSD With 2 clusters ...
Read now
Unlock full access