Links to 12-11-18 BARUG Presentations.Rmd

---
title: "Links to  12-11-18  BARUG Presentations"
author: "Joseph Rickert"
date: "12/12/2018"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```


### Agenda:
6:30 - Pizza and Networking
7:00 - Announcements
7:05 - Anirudh Acharya: An Introduction to the MXNet-R package (lightning talk)
7:20 - Emma Rudie - RStudio Connect at GRAIL
7:40 - Peter Li - Have you tesselated today?
8:05 - Dan Putler- Some Intuition on Why Collinearity Can Be "Bad", and When It Should Be a Concern (lightning talk)
8:20 - Rami Krispin - Introduction to the TSstudio package


### Links for Dan Putler's Talk
Some Intuition on Why Collinearity Can Be "Bad", and When It Should Be a Concern (lightning talk)

[PDF files and the R script for the Monte Carlo experiments](https://github.com/dputler/presentations/tree/master/BARUG_11_Dec_2018)


[The blog post my talk was based on](https://community.alteryx.com/t5/Data-Science-Blog/How-Concerned-Should-You-be-About-Predictor-Collinearity-It/ba-p/316701)


### Links to Anirudh Acharya's Talk

[MXNet R package](https://github.com/apache/incubator-mxnet/tree/master/R-package)
[MXNet R tutorials](http://mxnet.incubator.apache.org/tutorials/index.html#r-tutorials)
[MXNet R Documentation](https://s3.amazonaws.com/mxnet-prod/docs/R/mxnet-r-reference-manual.pdf)

The autoencoder example we went over will be posted here as a tutorial pretty soon)

### Emma Rudié Poem
**RStudio Connect**
R has always been, my go-to tool of choice
Statisticians far and wide, in R have found their voice
Commands which follow smoothly, from a math-y frame of mind
Almost any test you might conduct, has already been defined
Visualizing reproducibly is a cinch with ggplot
Rmarkdowns make it easy to show your train of thought
Plotly gives you hovertext, sliderbars, and more
And Shiny int’ractivity lets users independently explore
But as the language gives us more and more new ways to show our work
Sharing all these different files can leave one feeling irked
If ever you have wished for a single central hub
Like an internal equivalent to the wide-open RPubs
Then let me recommend to you: RStudio Connect
For we’ve recently adopted it, to resounding positive effect
Publishing is simple, with a single GUI click
Collaboration now has seen, a definite uptick
Choosing who sees what, and when reports should run
Is as easily as said as, in practice, done!
And if your ears want more, please patiently attend
For what here shall miss, this talk shall strive to mend