Tools for data analysis and modelling

Utrecht University offers and supports tools for data analysis and modelling.  

Introduction

In order to analyse data, model processes or run simulations you need two things: a computer (hardware) and applications or tools (software).

RDM support offers you services on high performance computing (HPC) and R Shiny facilities.

Choose your tool

I. Tools for interactive computing

What:
Utrecht University offers a range of desktop tools for data analysis, modelling an simulations: Rstudio/R, MATLAB, SPSS, SAS, STATA and many more. You may install these tools on your Solis desktop or laptop and some tools can also be installed on your private computer.

How:
Click here for an overview of available software and installation guidance. The licenses of some tools are restricted to researchers from specific faculties. Some may not be listed here.

II. High Performance Computing facilities

What:
A computer can be a workstation (desktop or laptop) or a high performance computing (HPC) facility. HPC gives you far more computing power, but comes with a price: you may have to adjust your scripts and must learn how to operate such a system. HPC computers don’t have a nice interactive, graphical user interface. Scripts must be submitted and scheduled via the command line. On an average workstation, parallelisation is limited to 2 or 3 threads or cores. On HPC you can use up to thousands. HPC also provides special computing devices (GPU) and larger memory (64GB+). Even if you don’t run your computations in parallel, you can use HPC to free up your workstation for other tasks.

How: 
Two HPC facilities are available for researchers of Utrecht University: 

On these systems the most used software tools are pre-installed, but it is also allowed to install your own software provide they are suited to run on the system and you have the appropriate license.

You can contact RDM Support for assistance to use the HPC facilities. 

III. Shiny webserver (R) to interact with and analyse data

What:
If you interactively want to showcase the results of your analysis, models and/or simulations you can submit apps to our Shiny webserver. Shiny webserver takes a fresh, interactive approach to telling your data story with R. 

Example:
See for example k-means clustering with the Iris dataset.

More information:
See Shiny from RStudio.

How:
You develop a Shiny app in Rstudio/R and upload this app to the webserver of Shiny.io. Everybody can see your app, but you need an access key to upload. 
You can apply for an access key at info.rdm@uu.nl.