R Shiny Application to Predict Survival on the Titanic

One of the most popular exercises to get started with Data Mining is predicting the survival on the Titanic. In the Kaggle website this is one of the main challenges, and you can find accurate documentation and tutorials on how to solve it using Excel, Python, R…

In the IBM Extreme Blue team that I leaded last summer the 4 students got started on Data Mining doing this challenge, and we end up creating a Shiny R application.  Shiny is a web application framework for R to convert your analysis into interactive web applications without having to write in HTML, CSS or JavaScript. We not only worked in solving the problem with R and with SPSS getting very good result, we created this application so everybody can access online and see the results. The model is created on the cloud and the results are calculated in as well in the cloud.

 http://titanicanalysis.mybluemix.net/

The application is hosted on Bluemix. To do so, we are using the Cloud Foundry R Custom Buildpack I modified to install R 3.1: https://github.com/aruizga7/cf-buildpack-r

You have the instructions about how to run R Shiny on Bluemix in this article: https://bluemixanalytics.wordpress.com/2014/09/11/new-developerworks-article-about-running-r-shiny-on-bluemix/

You can find the source code of this Shiny Application here: https://github.com/aruizga7/TitanicShinyApp.git

Continue reading

Special Event on Analytics for Bluemix

Next 18th of September I will be in Israel Tel Aviv running a special event about how to run Analytics on Bluemix. This event is part of the 200 IBM Bluemix Days. This event is quite special for two reasons, first because is the first time we focus on analytics on Bluemix and second because the workshop is going to take plage in the IBM Bluemix Garage of Tel Aviv.

- IBM Garage Alpha Zone Accelerator

-What is a Bluemix Garage?

200 IBM Bluemix Days

Continue reading

Buzz on Twitter about the death of Robin Williams using Bluemix and R

Today we got the sad news of Robin Williams death. Everybody is talking about it, and we are still getting a lot of last minute information, comments, emotions… in many different sites and of course social media.

To try to measure the buzz of this event on Twitter, I modified a Shiny application of Vivek Patil . The application is running in the IBM Bluemix platform. For this I the application is running R code to do the analysis. It is all deployed in IBM Bluemix and you can access here:

RandBluemix

startAppButton

 

Continue reading