People do all sorts of optimizations over their primary storage layer. AWS has built a whole product around a Redis based caching mechanism (to get sub-millisecond response times) https://aws.amazon.com/elasticache/redis/
[Read More]
R R6 classes and passing by reference
A simple test of passing by reference in R6 classes
By default, R6 classes are passed by reference. To avoid this, you need to use clone(deep = TRUE) while copying.
[Read More]
Machine learning for data management in pharma
AI on the way in, AI on the way out
This article is an excellent article about the data-management challenges faced by pharma companies. It outlines how GSK was faced with the self-realization that they are not equipped to make the best use of the vast quantities of digital data spewing out from their many clinical studies.
[Read More]
R, Travis and Github integration via GITHUB_PAT (personal access token)
Automating book publishing and collaboration
I started using RStudio BookDown to develop documentation for SciDB insight API.
[Read More]
A typical data science workflow
I found the following very interesting quote describing a typical data-science application workflow:
[Read More]