index.Rmd

---
title: "R Tools"
author: "Fabio & Simon "
date: "2018/05/17 <br> <br> Slides: <a href='r-user-group-stuttgart.github.io/18_17_05_altright/'>https://r-user-group-stuttgart.github.io/18_17_05_altright/</a>"
output:
  xaringan::moon_reader:
    lib_dir: libs
    nature:
      highlightStyle: github
      highlightLines: true
      countIncrementalSlides: false
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = F, warning = F, error = F, message = F)
options(htmltools.dir.version = FALSE)
```


## Outline

1. Social Media API(s)
2. Shiny App
3. Machine Learning
4. Make your own Slides in R


---
class: center, middle

## Get Started

---
class: inverse, center, middle

## Social Media Data API(s)


---

## APIS

+ [`SocialMediaLab`](https://github.com/vosonlab/SocialMediaLab)
+ [`tuber`: Access YouTube API via R](https://github.com/soodoku/tuber)
+ [`Rfacebook`: Access to Facebook API via R](https://github.com/pablobarbera/Rfacebook)
+ [`rtweet`](https://github.com/mkearney/rtweet)
+ [`instaR`: Access to Instagram API via R](https://github.com/pablobarbera/instaR)
+ [`4chan`: ChanStats](https://github.com/Vharuck/ChanStats)
+ [`tumblR`](https://github.com/klapaukh/tumblR)


---

## `rtweet`

```{r, echo = T}
library(rtweet)
library(dplyr)
rt <- search_tweets("#stuttgart", n = 500, include_rts = F)
glimpse(rt[,1:10])
```


---

```{r, echo = T, fig.align="center"}
gg1 <- ts_plot(rt, "1 hour") +
  ggplot2::theme_minimal() +
  ggplot2::labs(title = "Frequency of #stuttgart Twitter statuses from past 9 days")
gg1
```

---

```{r, echo = T, fig.align="center"}
library(plotly)
gg1 %>% ggplotly()
```


---

[Highcharter](http://jkunst.com/highcharter/)

```{r, echo = T, fig.align="center"}
library(highcharter)
rt %>% 
  mutate(h = created_at %>% lubridate::hour()) %>%
  group_by(h) %>% tally() %>%
  hchart("line", hcaes(h, n)) %>% hc_title(text = "Daily Hashtag Tweets #stuttgart")
```


---

class: inverse, center, middle

## Shiny App


---

## Shiny App


.pull-left[
* [Starting with Shiny](http://shiny.rstudio.com/tutorial/written-tutorial/lesson1/)
* [Building Shiny Apps Tutorial](https://deanattali.com/blog/building-shiny-apps-tutorial/)
* [Awesome Shiny Apps for Statistics](https://github.com/huyingjie/Awesome-shiny-apps-for-statistics)
]
.pull-right[
<img src = "http://hexb.in/hexagons/shiny.png" width = "200px">
]

---

class: inverse, center, middle

## Machine Learning


---


## Machine Learning Packages

* [caret](https://github.com/topepo/caret) 
    + 200 different classifiers
    + hyperparameter tuning by defult
* [h2o](https://github.com/h2oai/h2o-3/tree/master/h2o-r)
    + DNN
    + GBM
    + XGBOOST
* [keras](https://github.com/rstudio/keras)
    + Recurrent Neural Networks
    + Convolutional Neural Networks
    + LSTMs


---

## H2O

```{r, echo = T, eval = F}
fit_gbm <- h2o.gbm(
  x = x, 
  y = y, 
  training_frame = h2o_train_dtm, 
  distribution = "bernoulli"
)
```


---

class: inverse, center, middle

## Slides

---

## Slides

.pull-left[

* [xaringan Rmarkdown Slides](https://github.com/yihui/xaringan)
* Netlify
* Github
* Rmarkdown

]

.pull-right[

<img src = "https://img.ctolib.com/uploadImg/20170213/20170213192416_52.png" width = "200px">

]

---

class: inverse, center, middle


## Thanks for Listening!