R for ML

# R for ML
### Machine Learning with R <a href='https://therbootcamp.github.io'> Basel R Bootcamp </a> <a href='https://therbootcamp.github.io/ML_2019Oct/'> </a>  <a href='https://therbootcamp.github.io'> </a>  <a href='mailto:therbootcamp@gmail.com'> </a>  <a href='https://www.linkedin.com/company/basel-r-bootcamp/'> </a>
### October 2019

---

<div class="my-footer">
 
 
 <img src="https://raw.githubusercontent.com/therbootcamp/therbootcamp.github.io/master/_sessions/_image/by-sa.png" height=14 style="vertical-align: middle"/>
 
 <a href="https://therbootcamp.github.io/">
 
 
 www.therbootcamp.com
 
 
 </a>
 <a href="https://therbootcamp.github.io/">
 
 Machine Learning with R | October 2019
 
 </a>
 
 </div>

---

# R is a programming language

From [Wikipedia](https://en.wikipedia.org/wiki/Statistical_model) (emphasis added):

> A programming language is a **formal language** that specifies a set of instructions that can be used to produce various kinds of output. Programming languages generally consist of **instructions for a computer**. Programming languages can be used to create programs that **implement specific algorithms**.

```r
#data <- read.table(link)
#variables <- data[,c('group','variable')]
#analysis <- lm(variable ~ group, data = variables)
#summary(analysis)
```
]

---

# Why R?

Today, R is one of the **most popular languages for data science** and overall.

In terms of the number of data science jobs, **R beats SAS and Matlab**, and is on par with Python.

Image source: https://i0.wp.com/r4stats.com/

]

.pull-right6[

 <img src="https://i0.wp.com/r4stats.com/wp-content/uploads/2017/02/Fig-1a-IndeedJobs-2017.png" height="450">

]

---

# R is so popular because

There are many good reasons to prefer R over superficially more user friendly software such as **Excel** or **SPSS** or more complex programming languages like **C++** or **Python**.

.pull-left45[
### Pro
1. **It's free**
2. Relatively **easy**
3. **Extensibility** ([CRAN](https://cran.r-project.org/), packages)
4. **User base** (e.g., [stackoverflow](https://stackoverflow.com/))
5. [**Tidyverse**](https://www.tidyverse.org/) (`dplyr`, `ggplot`, etc.)
6. [**RStudio**](https://www.rstudio.com/)
7. **Productivity** options: [Latex](https://www.latex-project.org/), [Markdown](https://daringfireball.net/projects/markdown/), [GitHub](https://github.com/)
]

[Tidyverse](https://www.tidyverse.org/)
[Rcpp](http://www.rcpp.org/), [BH](https://cran.r-project.org/web/packages/BH/index.html): Links R to C++ and high-performance C++ libraries 
[rPython](http://rpython.r-forge.r-project.org/): Links R to Python 
[RHadoop](https://github.com/RevolutionAnalytics/RHadoop/wiki): Links R to Hadoop for big data applications. 
]

---

# R is great for ML

...because of high-performance R packages (extensions) downloaded and used millions of times.

---

# `caret`

The <high>C</high>lassification <high>A</high>nd <high>RE</high>gression <high>T</high>raining package is a meta-package to streamline the application of R's best machine learning tools.

`caret` facilitates... 
&nbsp;&nbsp;&nbsp;&nbsp;1) <high>data pre-processing</high> 
&nbsp;&nbsp;&nbsp;&nbsp;2) <high>feature selection</high> 
&nbsp;&nbsp;&nbsp;&nbsp;3) <high>fitting, tuning, & model prediction</high>

Includes dozens of algorithms/models including...

<high>regression</high>, <high>decision trees</high>, <high>random forests</high>, neural nets, AdaBoost, elastic nets, <high>lasso & ridge regression</high>, support vector machines, etc.

]

.pull-right4[
<img src="image/caret.png">
<img src="image/caret.png">
<img src="image/caret.png">
]
---

<div class="center_text">
 
 6+3+2 basic R lessons
 
</div>

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. <high>Everything is an object</high>
2. <high><mono><-</mono> creates/changes objects</high>
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

]

---

# 6+3+2 basic R lessons

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
# an object called one_two_three
one_two_three <- c(1, 2, 3)

# print object
one_two_three
```

```
## [1] 1 2 3
```

```r
# add 100 to the object's numbers (without <- )
one_two_three + 100
```

```
## [1] 101 102 103
```

```r
# print object (no <-, no change!)
one_two_three
```

```
## [1] 1 2 3
```
]

---

# 6+3+2 basic R lessons

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
# print object
one_two_three
```

```
## [1] 1 2 3
```

```r
# make change permanent (with <- )
one_two_three <- one_two_three + 100

# print object (it has changed!)
one_two_three
```

```
## [1] 101 102 103
```
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. <high>Everything happens through functions</high>
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
# function c()
one_two_three <- c(1, 2, 3)

# function `+`()
one_two_three + 100
```

```
## [1] 101 102 103
```

```r
# function print()
one_two_three
```

```
## [1] 1 2 3
```

```r
# function mean()
mean(x = one_two_three)
```

```
## [1] 2
```
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. <high>Functions have (default) arguments</high>
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
# no argument
mean()
```

```
## Error in mean.default(): argument "x" is missing, with no default
```

```r
# one (required) argument
mean(c(1, 2, 3))
```

```
## [1] 2
```

```r
# assume a missing value (NA)
mean(c(1, 2, 3, NA))
```

```
## [1] NA
```

```r
# changing default to handle NA
mean(c(1, 2, 3, NA), na.rm = TRUE)
```

```
## [1] 2
```
]

---

# 6+3+2 basic R lessons

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
# mean with pipe %>%
c(1, 2, 3) %>% mean()
```

```
## [1] 2
```

```r
# mean with pipe %>% and NA
c(1, 2, 3, NA) %>% mean()
```

```
## [1] NA
```

```r
# changing default to handle NA
c(1, 2, 3, NA) %>% mean(na.rm = TRUE)
```

```
## [1] 2
```
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. <high>Functions live in packages</high>
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

]

---

# 6+3+2 basic R lessons

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

<high>Install new packages</high> with `install.packages()`

```r
# install package: Only do this once!
install.packages("tidyverse")
```

<high>Load existing packages</high> with `library()`

```r
# load package: EVERY TIME you write code
library(tidyverse)
```

]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. <high>Find help with <mono>?</mono></high>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
?mean
```
<img src="https://raw.githubusercontent.com/therbootcamp/therbootcamp.github.io/master/_sessions/_image/help.png" width="500">
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. <high>Find help with <mono>?</mono></high>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
?cor
```
<img src="https://raw.githubusercontent.com/therbootcamp/therbootcamp.github.io/master/_sessions/_image/help_cor.png" width="500">
]

---

<h1><a>Interactive</a></h1>

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. <high>Data lives in data frames</high>
8. 3 data types + factors
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

.pull-right55[
<img src="https://raw.githubusercontent.com/therbootcamp/therbootcamp.github.io/master/_sessions/_image/main_objects.png"></img>
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. <high>3 data types + factors</high>
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

.pull-right5[
<img src="https://raw.githubusercontent.com/therbootcamp/therbootcamp.github.io/master/_sessions/_image/vector_types.png"></img>
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. <high>3 data types + factors</high>
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
print(baselers)
```

```
## # A tibble: 10,000 x 20
## id sex age height weight income
## <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
## 1 1 male 44 174. 113. 6300
## 2 2 male 65 180. 75.2 10900
## 3 3 fema… 31 168. 55.5 5100
## 4 4 male 27 209 93.8 4200
## 5 5 male 24 177. NA 4000
## 6 6 male 63 187. 67.4 11400
## 7 7 male 71 152. 83.3 12000
## 8 8 fema… 41 156. 67.8 7600
## 9 9 male 43 176. 69.3 8500
## 10 10 fema… 31 166. 66.3 6100
## # … with 9,990 more rows, and 14 more variables
```
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. <high>3 data types + factors</high>
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
# select sex veriable using $
baselers$sex
```

```
## [1] "male"   "male"   "female" "male"  
## [5] "male"   "male"   "male"   "female"
##  [ reached getOption("max.print") -- omitted 9992 entries ]
```

```r
# select sex veriable using %>% select
baselers %>% select(sex) %>% pull()
```

```
## [1] "male"   "male"   "female" "male"  
## [5] "male"   "male"   "male"   "female"
##  [ reached getOption("max.print") -- omitted 9992 entries ]
```

```r
# Possible, but less pretty...
baselers[['sex']]
baselers[[2]]
```

]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. <high>3 data types + factors</high>
9. `formula` and `data` specify a model

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
# original sex vector
baselers$sex
```

```
## [1] "male"   "male"   "female" "male"   "male"  
##  [ reached getOption("max.print") -- omitted 9995 entries ]
```

```r
as.factor(baselers$sex)
```

```
## [1] male   male   female male   male  
##  [ reached getOption("max.print") -- omitted 9995 entries ]
## Levels: female male
```

```r
as.factor(baselers$weight)
```

```
## [1] 113.4 75.2 55.5 93.8 <NA> 
## [ reached getOption("max.print") -- omitted 9995 entries ]
## 719 Levels: 37.9 38.3 39.2 39.6 40.3 ... 125.4
```

]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors</high>
9. <high><mono>formula</mono> and <mono>data</mono> specify a model</high>

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
# Run a regression and store result in my_lm
my_lm <- lm(formula = income ~ age + height,
 data = baselers)
```
 
<img src="https://raw.githubusercontent.com/therbootcamp/Erfurt_2018June/master/_sessions/_image/formula_description.png"></img>
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors</high>
9. <high><mono>formula</mono> and <mono>data</mono> specify a model</high>

Productivity essentials
10. Use RStudio and projects
11. Use editor, shortcuts, auto-complete
]

```r
# Include multiple terms with +
my_lm <- lm(formula = income ~ age + height,
 data = baselers)
```

<high>Include all variables</high> using `formula = y ~ .`

```r
# Use y ~ . to include ALL variables
my_lm <- lm(formula = income ~ .,
 data = baselers)
```

<high>Subtract variables</high> using `-`

```r
# Remove variables with -
my_lm <- lm(formula = income ~ . - id,
 data = baselers)
```

]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors</high>
9. <mono>formula</mono> and <mono>data</mono> specify a model

Productivity essentials
10. <high>Use RStudio and projects</high>
11. Use editor, shortcuts, auto-complete
]

.pull-right5[
### Projects help...
save workspace and history &#9679; set project specific options &#9679; access files &#9679; version control &#9679; etc.
<img src="https://raw.githubusercontent.com/therbootcamp/therbootcamp.github.io/master/_sessions/_image/projects_1.png" width="360">
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors</high>
9. <mono>formula</mono> and <mono>data</mono> specify a model

Productivity essentials
10. <high>Use RStudio and projects</high>
11. Use editor, shortcuts, auto-complete
]

.pull-right5[
### Folder structure
Complement projects by a <high>folder structure</high> appropriate for your project.
 

 <img src="https://raw.githubusercontent.com/therbootcamp/therbootcamp.github.io/master/_sessions/_image/folder_structure.png">

]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors</high>
9. <mono>formula</mono> and <mono>data</mono> specify a model

Productivity essentials
10. Use RStudio and projects
11. <high>Use editor, shortcuts, auto-complete</high>
]

<img src="https://raw.githubusercontent.com/therbootcamp/therbootcamp.github.io/master/_sessions/_image/editor_edit.png"></img>
 
Shortcut to <high>send to console</high>: <br2>&#8984;/ctrl + &#9166; <br2><br2>
Shortcut to <high>rerun chunk</high>: <br2> &#8984;/ctrl + &#8679; + p
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors</high>
9. <mono>formula</mono> and <mono>data</mono> specify a model

Productivity essentials
10. Use RStudio and projects
11. <high>Use editor, shortcuts, auto-complete</high>
]

```r
# Load packages with library()
library(tidyverse)
library(yarrr)
library(lme4)

# import data with 
baselers <- read_delim(file = "baselers.txt",
 delim = '\t')
```
]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors</high>
9. <mono>formula</mono> and <mono>data</mono> specify a model

Productivity essentials
10. Use RStudio and projects
11. <high>Use editor, shortcuts, auto-complete</high>
]

]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors</high>
9. <mono>formula</mono> and <mono>data</mono> specify a model

Productivity essentials
10. Use RStudio and projects
11. <high>Use editor, shortcuts, auto-complete</high>
]

]

---

# 6+3+2 basic R lessons

.pull-left4[
R essentials
1. Everything is an object
2. <mono><-</mono> creates/changes objects
3. Everything happens through functions
4. Functions have (default) arguments
5. Functions live in packages
6. Find help with <mono>?</mono>

Analytic essentials
7. Data lives in data frames
8. 3 data types + factors</high>
9. <mono>formula</mono> and <mono>data</mono> specify a model

Productivity essentials
10. Use RStudio and projects
11. <high>Use editor, shortcuts, auto-complete</high>
]

]

---

class: middle, center
<h1><a href="https://therbootcamp.github.io/ML_2019Oct/1_Data/baselrbootcamp_MLdata.zip">Download</a></h1>

---

<h1><a>Interactive</a></h1>