It’s been a while since the last update, but RSiteCatalyst is still going strong! Thanks to Wen for submitting a fix/enhancement to enable the ability to use multiple columns from a Classification within the QueueDataWarehouse function. No other bug fixes were made, nor was any additional functionality added.
Version 1.4.16 of RSiteCatalyst was submitted to CRAN yesterday and should be available for download in the coming days.
As I’ve mentioned in many a blog post before this one, I encourage all users of the software to continue reporting bugs via GitHub issues, and especially if you can provide a working code example. Even better, a fix via pull request will ensure that your bug will be addressed in a timely manner and for the benefit to others in the community.
Note: Please don’t email directly via the email in the RSiteCatalyst package, it will not be returned. Having a valid email contact in the package is a requirement to have a package listed on CRAN so they can contact the package author, it is not meant to imply I can/will provide endless, personalized support for free.
The talk outlines using the UDP listener for StreamSets to collect packets from the F1 2018 game, writing the packets to Kafka, reading from Kafka and using Groovy to parse the packets, and using the OmniSci JDBC driver to insert the data into one of nine OmniSciDB tables. With this workflow, you have a robust platform for accelerated analytics, using the power of GPUs for fast computation.
In this webinar sponsored by the Open Data Science Conference (ODSC), I outline a brief history of GPU analytics and the problems that using GPU analytics solves relative to using other parallel computation methods such as Hadoop. I also demonstrate how OmniSci fits into the broader GPU-accelerated data science workflow, with examples provided using Python.
Check out the video, grab the Jupyter Notebook from the odscwebinar repo and get started with OmniSci and GPU-accelerated data science!