Introduction to Statistical Computing on the Flux Cluster, April 9th, 3:30-5:00 pm

Tuesday, April 9, 2013 3:30-5:30 pm Room 2001A – Modern Languages Building David Childers, CSCAR and Bennet Fauber, CAEN HPC/Flux This workshop assumes no prior knowledge of Flux or parallel computing. It may be useful for anyone who anticipates having many, repetitive statistical analyses, or who will be using highly iterative methods such as Monte Carlo simulation, bootstrapping/resampling, and some kinds of missing data imputation. This workshop will provide a basic introduction to batch processing on the Flux computing cluster, including how to connect to Flux and copy files to and from it, a brief introduction to PBS (Portable Batch System), writing a basic PBS script, and submitting a job. Once that is complete, we will show examples of running SAS, Stata, and R on the cluster. We will then give a very brief introduction to parallel computing with R. Updated March 26th: this workshop is now filled to capacity.  A wait-list is being created for those who register after 3/26 in the event that cancellations are received. If enough interested people register on the wait-list, we will look into scheduling a second workshop in the near future. Please register for the workshop (wait-list) by Wednesday, April 4th. There is no cost to attend. You will also need to obtain an MToken and get an account on Flux at least two days prior to the workshop. Information about obtaining an MToken can be found at http://cac.engin.umich.edu/resources/loginnodes/twofactor.html and the account application form is at https://www.engin.umich.edu/form/cacaccountapplication. If you have questions about the workshop, please send e-mail to statistics-on-flux@umich.edu This workshop is co-sponsored by the Center for Statistical Consultation and Research (CSCAR) and the Office of Research Cyberinfrastructure (ORCI).