IBM Data Science Experience Desktop could be your very own data science lab in your basement.



IBM has recently released a cool new ‘Desktop’ version of Data Science Experience (DSX) https://ibm.co/2pQEK2h.

What does it do?

  1. It makes the data science experience ‘up close and personal’ on your desktop, meaning you can run it right on your PC! I know, I was a bit skeptical at first. But, it will be a great way for someone who wants to learn and harness their data science skills without having to be connected to the internet. There are times I wish I had DSX on my desktop, especially when I was traveling or working away from office where data connectivity was sometimes not available or slow.
  2. It provides you with the access to the tutorials, and materials from DSX community that really helps you to start learning the technologies involved with data science experience. I believe the Desktop version is useful for those who would like to get in touch with many different technologies (such as Spark, SparkML, R, Jupyter, etc.) but don’t know where or how to start. The installation of DSX Desktop is very straight forward, and you will be able to start using those the moment you finished installing it on your desktop. I mention again that tutorials and materials for you to start exploring is all there right from the start.
  3. Even though the DSX Desktop is limited in such a way that you will only be able to use local files (Yes, the Desktop wouldn’t let you add a data asset that’s out on the cloud or in your remote database. Some might find this a bit disappointing), its functions are on par with the cloud version, and functionality wise, you are only restricted on your desktop’s computational capability.

DSX Desktop is both PC and Mac compatible. Installing the Desktop can be downloaded and will require you to also install a Docker engine on your computer. Those who are not familiar with the ‘Docker,’ it is a form of a virtualization technology that lets you run multiple applications (inside the ‘containers’) without having multiple instances of the virtual machine and operating system, but rather sharing much of the kernel and  resources of your host OS. More details about how docker works is described here: http://dockr.ly/2q9ODuQ

What I noticed when I was using the Docker on windows, is that it actually requires an Oracle VM (virtualbox) to make it run. In the most recent version of the docker, this restriction is removed, as it runs on top of windows 10 and mac operating system natively. However, with older versions of windows (like myself), you have to run the docker on top of a virtual linux kernel that runs on top of the operating system. That’s why you are seeing the ‘oracle VM virtualbox’ being installed when you are installing the docker toolbox (so, in a sense this is not a native docker environment).

After installing and using the DSX Desktop for a couple of days (and it may be just too early to say much about it), I am beginning to like it more and more. It is a lot more convenient to quickly double click the button on your desktop, and start browsing through the community samples and tutorials, and immediately start toying around with it. If I have a CSV file lying around my desktop, I can easily import it and use Jupyter notebook and run python scripts to convert it into various visualizations. I can also try learn how to use R studio conveniently within the desktop, not to mention exploring the world of machine learning without having to install any of it, or access anywhere.

It just felt like I am suddenly a in-house scientist experimenting many lab experiments in my basement. I must confess it felt pretty awesome.

 

Share this post:






#Analytics,#bluemix,#awvi,#hybridcloud,#ibm

Bluemix

via Bluemix Blog https://ibm.co/2pQcNaA

May 5, 2017 at 02:24AM