IBM DSx Experience
January 30, 2018

IBM DSx Experience

Anonymous | TrustRadius Reviewer
Score 6 out of 10
Vetted Review
Verified User

Overall Satisfaction with IBM Data Science Experience (DSx)

DSx was used as a test environment in my organization for applying advanced analytical techniques like machine learning, parallel processing etc. It was used by the analytics department solely and not across the whole organization.
  • I particularly like working on R for ML problems. DSx provides both Jupyter notebook and R studio interfaces for doing the same. Which is fantastic in terms of flexibility and applicability.
  • There are multiple used cases explained in the community section so that one can learn and apply the knowledge at the same time.
  • Ease of navigation was of a fantastic magnitude using the easy drop down menus.
  • Apache spark connection to R Studio tool keeps on disconnecting. Lot of room for improvement there. A stable connection helps the user have a good experience.
  • Many ML functionalities under H2O package in R don't seem to work in the Apache Spark environment.
  • If a documentation could be provided regarding ML using Apache Spark in R that would be really helpful.
  • Knowledge gaining.
  • Alternative environment exploration.
  • Exposure handling.
To my knowledge, DSx is very well suited for handling any data size using pySpark. However, it is not very well suited for using sparklyr since many of the functionalities doesn't work as it should. Less availability of documentation makes things worse. One could use Java and Scala as well for Apache Spark. But these languages are not so famous among analysts.