addressalign-toparrow-leftarrow-rightbackbellblockcalendarcameraccwcheckchevron-downchevron-leftchevron-rightchevron-small-downchevron-small-leftchevron-small-rightchevron-small-upchevron-upcircle-with-checkcircle-with-crosscircle-with-pluscrossdots-three-verticaleditemptyheartexporteye-with-lineeyefacebookfolderfullheartglobegmailgooglegroupshelp-with-circleimageimagesinstagramlinklocation-pinm-swarmSearchmailmessagesminusmoremuplabelShape 3 + Rectangle 1ShapeoutlookpersonJoin Group on CardStartprice-ribbonShapeShapeShapeShapeImported LayersImported LayersImported Layersshieldstartickettrashtriangle-downtriangle-uptwitteruserwarningyahoo

"Official"­ August 2014 Meetup

  • Aug 12, 2014 · 6:30 PM
  • A9

6:30 - Networking and Pizza
7:00 - Announcements
7:05 - Spencer Graves - Animation with R
7:35 - Norm Matloff - An R Package for parallel Matrix Powers
8:10 - Joseph Rickert - GLMs on large data sets

Abstract - Animation with R

This presentation will describe an "Animate" function in the "Ecfun" package on R-Forge.  This function provides a way to create large numbers of image files (in, e.g., png at 25, 30 or 60 frames per second) that can then be mated with audio to produce a movie.  The appearance of motion can be obtained with any capability for producing the requisite number of images.  The "animate" function makes it easier in some cases to produce various effects.  Examples include a series of plots that add points or change colors as time progresses.  This can be used to remix images created elsewhere, e.g., panning and zooming on a photograph or a map can create the illusion of motion.  It can be used with other video software such as FFmpeg that supports merging image files with audio.  The presentation will also discuss general concepts for functional programming such as parsing, modifying and executing variations of a function.  It will also discuss concepts of defensive programming to produce diagnostics to help a user find and fix problems.


Abstract - An R Package for Parallel Matrix Powers

Powers of square matrices are useful in a variety of contexts. Two examples in the statistics/data sciencefield are calculating the stationary distribution of a discrete-time Markov chain and determining whether agraph is connected. Also, the exponential of a matrix M, defined as eM =P∞r=0 Mr/r!, can be used tocalculate the finite-time distribution of a continuous-time Markov chain (Stewart, 2009).The CRAN package expm calculates matrix exponentials and powers. However, in many applications, thematrices involved are quite large, so parallel computation is needed. Our package parmatpows ( fills this need. Two parallel platforms are supported,multicore and GPU, based on the CRAN packages Rdsm and gmatrix, affording excellent speedups.


Abstract - GLMs on large data sets

Since their introduction by Nelder and Wedderburn over forty years ago, Generalized Linear Models (GLMs) have been a mainstay of statistical inference.  Moreover, while they were originally employed with samples of relatively modest size, GLMs are now being employed with very large data sets in machine learning and data science applications. In this talk, we will briefly review the history of using GLMs in R, discuss the issues involved in using GLMs on large data sets, and show examples of various models including logistic regression, Poisson and Tweedie models running on large data sets using the Parallel External Memory algorithms implemented in Revolution R Enterprise.

Join or login to comment.

  • Joseph R.

    Many thanks to the folks at A9 for their hospitality and to RStudio for the pizza.

    August 13, 2014

  • Jerry C.

    I really liked the format of this event. Each speaker presented their material for about 20 minutes, followed by questions from the audience. It was a great idea having such diverse topics. It really seemed to keep the audience engaged.

    August 13, 2014

  • Jian J. S.

    Pretty good presentation

    August 12, 2014

  • Jen B.

    Sorry to miss this, conflicts with another meeting.

    August 12, 2014

  • A former member
    A former member

    Anyone up for doing some data science/machine learning projects in R and/or Python on a contract basis? We're working on some financial models, and we could use some help.

    August 12, 2014

    • Dennis N.

      I'm interested in talking with you about this. I have R experience, MA in Statistics.

      August 12, 2014

    • Rahul C.

      I am up for it. R programmer with experience in Financial Risk Modeling, Backtesting Investment strategies, Corp Finance models..

      August 12, 2014

  • Jerry C.

    This is the first Meetup that I'll be attending. I'm really looking forward to meeting the group and learning more about R. It should be fun!

    August 12, 2014

  • Andrew Defries P.

    R U ready?

    August 11, 2014

  • Daniel F.

    Enjoy a wonderful evening in PA. :-)

    August 11, 2014

Our Sponsors

  • RStudio

    Financial support for meetings

  • O'Reilly

    Our March 29th BARUG meeting onsite at Strata Hadoop World

People in this
Meetup are also in:

Sign up

Meetup members, Log in

By clicking "Sign up" or "Sign up using Facebook", you confirm that you accept our Terms of Service & Privacy Policy