Introduction to R

### Maximilian H.K. Hesselbarth

#### University of Michigan (EEB)

2022-06-30 / 2022-07-01 
]

.pull-right[
<img src="data:image/png;base64,#img/first-then.png" width="85%" style="display: block; margin: auto;" />

---

All slides available at: [https://mhesselbarth.github.io/introduction-r-workshop/](https://mhesselbarth.github.io/introduction-r-workshop/)

---

# About myself

.pull-left[
<img src="data:image/png;base64,#img/hex-logos.png" width="100%" style="display: block; margin: auto;" />
]

.pull-right[
- Member of the [Coastal Ecology and Conservation Lab](https://www.jacoballgeier.com) .small[("Allgeier Lab")]

- Working on **individual-based simulation modelling**, **landscape ecology**, **point pattern analysis** and spatial data in general

- Author/Contributor of several _R_ packages: _landscapemetrics_ .ref[(Hesselbarth et al. 2019)], _shar_ .ref[(Hesselbarth 2021)], _arrR_ .ref[(Esquivel et al. 2022)] and others
]

---

## Section 1: Basic introduction

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# _R_ programming language

- Widely used programming language for **statistical analysis**, **data science** and much more

- [_R_](https://www.r-project.org) is **free**, **open-source**, and **multi-platform** (_Windows_, _macOS_, _Linux_)

- Allows open, reproducible, and transparent research

- Very active and generally **friendly community** .small[(<svg viewBox="0 0 512 512" style="height:1em;position:relative;display:inline-block;top:.1em;" xmlns="http://www.w3.org/2000/svg">  <path d="M459.37 151.716c.325 4.548.325 9.097.325 13.645 0 138.72-105.583 298.558-298.558 298.558-59.452 0-114.68-17.219-161.137-47.106 8.447.974 16.568 1.299 25.34 1.299 49.055 0 94.213-16.568 130.274-44.832-46.132-.975-84.792-31.188-98.112-72.772 6.498.974 12.995 1.624 19.818 1.624 9.421 0 18.843-1.3 27.614-3.573-48.081-9.747-84.143-51.98-84.143-102.985v-1.299c13.969 7.797 30.214 12.67 47.431 13.319-28.264-18.843-46.781-51.005-46.781-87.391 0-19.492 5.197-37.36 14.294-52.954 51.655 63.675 129.3 105.258 216.365 109.807-1.624-7.797-2.599-15.918-2.599-24.04 0-57.828 46.782-104.934 104.934-104.934 30.213 0 57.502 12.67 76.67 33.137 23.715-4.548 46.456-13.32 66.599-25.34-7.798 24.366-24.366 44.833-46.132 57.827 21.117-2.273 41.584-8.122 60.426-16.243-14.292 20.791-32.161 39.308-52.628 54.253z"></path></svg> [#rstats](https://twitter.com/hashtag/RStats), [#rspatial](https://twitter.com/hashtag/RSpatial))]

- Very popular in ecology .ref[(Atkins et al. 2022, Hesselbarth et al. 2021, Joo et al. 2020, Lai et al. 2019)]

- Integration of other programming languages (e.g., `Python`, `C`, `C++`)

---

background-image: url("data:image/png;base64,#img/rstudio-logo.png")
background-position: 95% 5%
background-size: 15%

# _RStudio_

- [_RStudio_](https://www.rstudio.com) as **integrated development environment** (IDE)

- **Write**, save and **run** _R_ code scripts

- Includes syntax-highlighting, auto-completion, figure and help panels, ...

- Allows to use _RStudio_ projects (_.Rproj_) to organize related scripts/figure/data ...

---

background-image: url("data:image/png;base64,#img/rstudio-logo.png")
background-position: 95% 5%
background-size: 15%

# ...some advice...

.pull-left[
<img src="data:image/png;base64,#img/set-fire-rm.png" width=" 100%" style="display: block; margin: auto;" />
]

.pull-right[
<img src="data:image/png;base64,#img/workspace.png" width=" 100%" style="display: block; margin: auto;" />
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# _R_ as a calculator

`-` Subtraction

`*` Multiplication

`/` Division

`^` Exponentiation
]

```r
# add two numbers
14 + 5
## [1] 19

# combination of different operators
(23 - 5) / (4 * 5) ^ 2
## [1] 0.045
```

`#` for comments in code
]

<br>

`==`, `!=`, `>=`, `<=` Logical operators

`&` "and" statement

`|` "or" statement

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# _R_ Objects

- Once objects are assigned using the `<-` operator, they can be **reused**

- Use `snake_case` or `camelCase` naming 
]

```r
# that's my age
my_age <- 31

# that's the global life expectancy
global_expectancy <- 90

# ...well...
*(progress <- (my_age / global_expectancy) * 100)
## [1] 34.44444

# check if I'm still below 1/2
progress <= 50
## [1] TRUE
```
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

- integer : `5L`

- character : `"fish"`

- logical : `TRUE`/`FALSE`

- .grey[(complex : `1+4i`)]
]

- `vector c()`
Collection of elements/values of same type

- `matrix()`
Multi-dimensional _vector_ (rows, columns)

- `data.frame()`
Tabular data that allows different types in columns, but same number of rows

- `list()`
List of different types and/or structures (allows all of above)
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Vectorization

- Vectorization one major **strength** of _R_

- Operations are applied to **all** elements of vector (or combinations)

```r
# create named vector with extinction numbers
extinct <- c("amphibians" = 35, "birds" = 159, "fish" = 80, "mammals"  = 85)

# log of all values
log(extinct)
## amphibians      birds       fish    mammals 
##   3.555348   5.068904   4.382027   4.442651

# calc relative number in relation to maximum
(max_number <- max(extinct))
## [1] 159
*extinct / max_number * 100
## amphibians      birds       fish    mammals 
##   22.01258  100.00000   50.31447   53.45912

# multiply all values with log values
*extinct * log(extinct)
## amphibians      birds       fish    mammals 
##   124.4372   805.9558   350.5621   377.6254
```

---

# Palmer penguins dataset

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Indexing .tiny[(1/3)]

```r
# create species vector (characters)
species <- c("Adelie", "Chinstrap", "Gentoo")

class(species)
## [1] "character"

# subset second object
*species[2]
## [1] "Chinstrap"

# subset first and third object
*species[c(1, 3)]
## [1] "Adelie" "Gentoo"

# subset everything BUT second object
*species[-2]
## [1] "Adelie" "Gentoo"

(species[1:3] <- c("spec_1", "spec_2", "spec_3"))
## [1] "spec_1" "spec_2" "spec_3"

c(species, "unknown")
## [1] "spec_1"  "spec_2"  "spec_3"  "unknown"
```
]

- **Positive** or **negative** indexing possible
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Indexing .tiny[(2/3)]

- _data.frame_ columns and **named elements** can be accessed using `object$name`
]

```r
class(penguins)
## [1] "tbl_df"     "tbl"        "data.frame"

# subset rows 1-5 and columns 1, 3, 8
*penguins[1:5, c(1, 3, 8)]
## # A tibble: 5 × 3
##   species bill_length_mm  year
##   <fct>            <dbl> <int>
## 1 Adelie            39.1  2007
## 2 Adelie            39.5  2007
## 3 Adelie            40.3  2007
## 4 Adelie            NA    2007
## 5 Adelie            36.7  2007

# subset body mass column as vector
body_mass <- penguins$body_mass_g
```
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Indexing .tiny[(3/3)]

- Logical **tests** can be used to subset using `object[TRUE/FALSE-vector]`

- If either row or column index is empty, **all** are returned

```r
# subset all individuals with body mass larger than 4000
*penguins[penguins$body_mass_g > 4000, ]
## # A tibble: 174 × 8
##    species island    bill_length_mm bill_depth_mm flipper_length_mm body_mass_g
##    <fct>   <fct>              <dbl>         <dbl>             <int>       <int>
##  1 <NA>    <NA>                NA            NA                  NA          NA
##  2 Adelie  Torgersen           39.2          19.6               195        4675
##  3 Adelie  Torgersen           42            20.2               190        4250
##  4 Adelie  Torgersen           34.6          21.1               198        4400
##  5 Adelie  Torgersen           42.5          20.7               197        4500
##  6 Adelie  Torgersen           46            21.5               194        4200
##  7 Adelie  Dream               39.2          21.1               196        4150
##  8 Adelie  Dream               39.8          19.1               184        4650
##  9 Adelie  Dream               44.1          19.7               196        4400
## 10 Adelie  Dream               39.6          18.8               190        4600
## # … with 164 more rows, and 2 more variables: sex <fct>, year <int>
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Functions .tiny[(1/3)]

- **Packages** are **collections** of functions provided by the community

```r
library(palmerpenguins)
library(scales)
```
]

- Functions are evaluated from **inside** to **outside**

- Pass values by name or argument

```r
# standardize body mass to 0-1 scale, 
# calculate mean, and calculate log
log(mean(rescale(body_mass, to = c(0, 1)), na.rm = TRUE))
## [1] -0.8742998

head(rescale(body_mass, to = c(1, 2)))
## [1] 1.291667 1.305556 1.152778       NA 1.208333 1.263889
head(rescale(body_mass, c(1, 2)))
## [1] 1.291667 1.305556 1.152778       NA 1.208333 1.263889
```
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Functions .tiny[(2/3)]

- **Error messages** often give hints whats wrong

```r
min(penguins$sex)
## Error in Summary.factor(structure(c(2L, 1L, 1L, NA, 1L, 2L, 1L, 2L, NA, : 'min' not meaningful for factors
```

- Read and decided what to do with **warning messages**

```r
mean(penguins$sex)
## Warning in mean.default(penguins$sex): argument is not numeric or logical:
## returning NA
## [1] NA
```

- Use e.g., `?function_name` to open **documentation** for help (or highlight + F1)
]

.pull-right[
<img src="data:image/png;base64,#img/help-page.png" width="100%" style="display: block; margin: auto;" />
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Functions .tiny[(3/3)]

- Very easy to create **own** functions using `fun_name <- function(arg1, arg2,...){body}`

- _R_ always returns the **last line** or explicit `return()` statement.

```r
*calc_area <- function(x, y, total = FALSE, na.rm = TRUE){
  
  # multiply length * width
  area <- x * y
  
  # calculate sum of all values
  if (total) {
*   area <- sum(area, na.rm = na.rm)
  }
  
* return(area)
}

head(calc_area(penguins$bill_length_mm, penguins$bill_depth_mm))
## [1] 731.17 687.30 725.40     NA 708.31 809.58

calc_area(penguins$bill_length_mm, penguins$bill_depth_mm, total = TRUE)
## [1] 256768.7
```

---

background-image: url("data:image/png;base64,#img/meme-google.png")
background-position: 50% 25%
background-size: 35%

## Exercise 1: Basic introduction

... work on [_exercise-section-1.Rmd_](https://github.com/mhesselbarth/introduction-r-workshop/blob/main/exercises/exercise-section-1.Rmd) ...

---

## Section 2: Data wrangling and some stats

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Importing data

- `read.table()` to import tabular **text files** (e.g., _file.csv_)

- _readxl_ packages to import **Excel** files

- `getwd()`/`setwd()` for current working directory (or better `here()` frome _here_ package)

```r
# import tabular text data file (most robust)
penguins_csv <- read.table(file = "data/penguins.csv", 
*                          header = TRUE, sep = ",")

# read MS Excel file
penguins_excel <- readxl::read_xlsx(path = "data/penguins.xlsx", sheet = "clean", 
                                    col_types = c("text", "text", "numeric", "numeric",
                                                  "numeric", "numeric", "guess", "guess"))

head(penguins_excel, n = 3)
## # A tibble: 3 × 8
##   species island bill_length_mm bill_depth_mm flipper_length_… body_mass_g sex  
##   <chr>   <chr>           <dbl>         <dbl>            <dbl>       <dbl> <chr>
## 1 Adelie  Torge…           39.1          18.7              181        3750 male 
## 2 Adelie  Torge…           39.5          17.4              186        3800 fema…
## 3 Adelie  Torge…           40.3          18                195        3250 fema…
## # … with 1 more variable: year <dbl>
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Modify _data.frames_

```r
# remove all rows that have NA value
head(complete.cases(penguins))
## [1]  TRUE  TRUE  TRUE FALSE  TRUE  TRUE
penguins <- penguins[complete.cases(penguins), ]

# modify existing columns
*penguins$year <- factor(penguins$year)

# create new columns based on existing ones
penguins$body_mass_kg <- penguins$body_mass_g / 
  10000

penguins$bill_area <- penguins$bill_length_mm * 
  penguins$bill_depth_mm

# completely new
penguins$rand <- runif(n = nrow(penguins))
```
]

---

background-image: url("data:image/png;base64,#img/dplyr.png")
background-position: 95% 5%
background-size: 10%

# _dplyr_ package .tiny[(1/3)]

- Many useful functions (mostly) to deal with _data.frames_

- Also provides **pipe** operator, which allows to write `x %>% f() %>% g()` instead of `g(f(x))`
]

```r
# filter by sex and body mass
*dplyr::filter(penguins, sex == "female",
*             body_mass_kg > 0.5) %>%
  # return only selected columns
  dplyr::select(species, sex, body_mass_kg) %>% 
  # show last 5 rows
  tail(n = 5)
## # A tibble: 5 × 3
##   species sex    body_mass_kg
##   <fct>   <fct>         <dbl>
## 1 Gentoo  female        0.505
## 2 Gentoo  female        0.515
## 3 Gentoo  female        0.51 
## 4 Gentoo  female        0.52 
## 5 Gentoo  female        0.52
```
]

---

background-image: url("data:image/png;base64,#img/dplyr.png")
background-position: 95% 5%
background-size: 10%

# _dplyr_ package .tiny[(2/3)]

- `mutate()` to modify or add _data.frame_ columns

- Can be combined with `case_when()` for if-else statements

```r
dplyr::filter(penguins, sex == "female", body_mass_kg > 0.5) %>% 
  dplyr::select(species, sex, body_mass_kg) %>% 
* dplyr::mutate(body_mass_pounds = body_mass_kg * 2.20462,
*               body_class = dplyr::case_when(body_mass_pounds < 1.125 ~ "light",
*                                             body_mass_pounds > 1.125 ~ "heavy"))
## # A tibble: 5 × 5
##   species sex    body_mass_kg body_mass_pounds body_class
##   <fct>   <fct>         <dbl>            <dbl> <chr>     
## 1 Gentoo  female        0.505             1.11 light     
## 2 Gentoo  female        0.515             1.14 heavy     
## 3 Gentoo  female        0.51              1.12 light     
## 4 Gentoo  female        0.52              1.15 heavy     
## 5 Gentoo  female        0.52              1.15 heavy
```

---

background-image: url("data:image/png;base64,#img/dplyr.png")
background-position: 95% 5%
background-size: 10%

# _dplyr_ package .tiny[(3/3)]

```r
dplyr::group_by(penguins, species) %>%
  dplyr::summarise(n = dplyr::n(), 
                   max_g = max(body_mass_g))
## # A tibble: 3 × 3
##   species       n max_g
##   <fct>     <int> <int>
## 1 Adelie      146  4775
## 2 Chinstrap    68  4800
## 3 Gentoo      119  6300
```
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Descriptive stats

- Many **descriptive** stats e.g., `min()`, `max()`, `mean()`, `sd()`, `range()`, `quantile()`, `density()`, etc.

- `summary()` very powerful function that returns **different outputs** based on object

```r
summary(penguins[, -c(9, 10, 11)])
##       species          island    bill_length_mm  bill_depth_mm  
##  Adelie   :146   Biscoe   :163   Min.   :32.10   Min.   :13.10  
##  Chinstrap: 68   Dream    :123   1st Qu.:39.50   1st Qu.:15.60  
##  Gentoo   :119   Torgersen: 47   Median :44.50   Median :17.30  
##                                  Mean   :43.99   Mean   :17.16  
##                                  3rd Qu.:48.60   3rd Qu.:18.70  
##                                  Max.   :59.60   Max.   :21.50  
##  flipper_length_mm  body_mass_g       sex        year    
##  Min.   :172       Min.   :2700   female:165   2007:103  
##  1st Qu.:190       1st Qu.:3550   male  :168   2008:113  
##  Median :197       Median :4050                2009:117  
##  Mean   :201       Mean   :4207                          
##  3rd Qu.:213       3rd Qu.:4775                          
##  Max.   :231       Max.   :6300
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Correlation

- Calculate and test for **correlation** between two _vectors_ using `cor()`/`cor.test()`

```r
# calculate correlation between body mass and bill length
*cor(x = penguins$body_mass_g, y = penguins$bill_length_mm)
## [1] 0.5894511

# correlation test using formula syntax
*cor.test(penguins$body_mass_g, penguins$bill_length_mm)
## 
## 	Pearson's product-moment correlation
## 
## data:  penguins$body_mass_g and penguins$bill_length_mm
## t = 13.276, df = 331, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.5145745 0.6554058
## sample estimates:
##       cor 
## 0.5894511
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Student's t-Test

- Student's t-Test to test if the **means** of two groups are **different**

- Check documentation for **additional arguments** and test specifications

```r
# extract male and femalae body mass values
male <- penguins[penguins$sex == "male" & penguins$species == "Adelie", "body_mass_g", drop = TRUE]
female <- penguins[penguins$sex == "female" & penguins$species == "Adelie", "body_mass_g", drop = TRUE]

# compute t.test between male and female
*t.test(male, female, conf.level = 0.95)
## 
## 	Welch Two Sample t-test
## 
## data:  male and female
## t = 13.126, df = 135.69, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  573.0139 776.3012
## sample estimates:
## mean of x mean of y 
##  4043.493  3368.836

## same but a bit more elegant use formula syntax
*# t.test(body_mass_g ~ sex, data = penguins, subset = species == "Adelie")
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

### Linear regression .tiny[(1/2)]

- `lm()` to fit **linear regression** model using formula syntax

- `summary()` behaves different for _lm_ than for _data.frame_ object

```r
# fit linear model with variables bill length (dependent) and body mass (independent)
*mass_length_lm <- lm(formula = bill_length_mm ~ body_mass_g, data = penguins)

# get coefficients
summary(mass_length_lm)
## 
## Call:
## lm(formula = bill_length_mm ~ body_mass_g, data = penguins)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -10.1652  -3.0664  -0.7672   2.2356  16.0371 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 2.715e+01  1.292e+00   21.02   <2e-16 ***
## body_mass_g 4.003e-03  3.016e-04   13.28   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.424 on 331 degrees of freedom
## Multiple R-squared:  0.3475,	Adjusted R-squared:  0.3455 
## F-statistic: 176.2 on 1 and 331 DF,  p-value: < 2.2e-16
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

### Linear regression .tiny[(2/2)]

- `predict()` allows to predict dependent variable based on new independent values

```r
bill_predictions <- predict(mass_length_lm, 
*                           newdata = data.frame(body_mass_g = seq(from = 3000, to = 6000, by = 250)))

bill_predictions
##        1        2        3        4        5        6        7        8 
## 39.16059 40.16142 41.16224 42.16306 43.16388 44.16471 45.16553 46.16635 
##        9       10       11       12       13 
## 47.16717 48.16800 49.16882 50.16964 51.17046
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# ANOVA

- ANOVA to test if independent variable(s) affect numerical dependent variable (**more than 2 groups**)

- Use `*` instead of `+` for **interaction** term

```r
# test of flipper length depends on sex (2 levels) and species (3 levels)
flipper_anova <- aov(flipper_length_mm ~ sex + species, data = penguins)

# check anova results
*summary(flipper_anova)
##              Df Sum Sq Mean Sq F value Pr(>F)    
## sex           1   4246    4246   129.5 <2e-16 ***
## species       2  50185   25093   765.3 <2e-16 ***
## Residuals   329  10787      33                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Tukey's post-hoc test

- **Post-hoc test** to see difference between groups of independent variable levels

```r
TukeyHSD(flipper_anova)
##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = flipper_length_mm ~ sex + species, data = penguins)
## 
## $sex
##                 diff      lwr      upr p adj
## male-female 7.142316 5.907707 8.376925     0
## 
## $species
##                      diff       lwr       upr p adj
## Chinstrap-Adelie  5.72079  3.741515  7.700065     0
## Gentoo-Adelie    27.04253 25.377568 28.707483     0
## Gentoo-Chinstrap 21.32174 19.272353 23.371118     0
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Assumptions .tiny[(1/2)]

- Independence

- Normality (`shapiro.test()`)

- Homogeneity of variances (`var.test()`)

- Random Sampling

]

- Independence

- Normality (`shapiro.test()`)

- Homoscedasticity (`plot(lm_object)`)

- Linearity (`plot(y ~ x)`)
]

```r
*shapiro.test(male)
## 
## 	Shapiro-Wilk normality test
## 
## data:  male
## W = 0.98269, p-value = 0.416
*shapiro.test(female)
## 
## 	Shapiro-Wilk normality test
## 
## data:  female
## W = 0.97684, p-value = 0.1985
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Assumptions .tiny[(2/2)]

- `plot(mass_length_lm)` returns four plots helping with model assumptions

---

background-image: url("data:image/png;base64,#img/lm.png")
background-position: 50% 35%
background-size: 50%

## Exercise 2: Basic stats

... work on [_exercise-section-2.Rmd_](https://github.com/mhesselbarth/introduction-r-workshop/blob/main/exercises/exercise-section-2.Rmd) ...

---

## Section 3: Data vizualisation

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# _base_ vs. _ggplot2_

- _R_ has two major visualization environments: _base_ and _ggplot2_

- Today, we are going to focus on _base_

.pull-left[
<div class="figure" style="text-align: center">
<img src="data:image/png;base64,#index_files/figure-html/base_plot-1.png" alt="base plot" width="80%" />
<p class="caption">base plot</p>
</div>
]

.pull-right[
<div class="figure" style="text-align: center">
<img src="data:image/png;base64,#index_files/figure-html/ggplot-1.png" alt="ggplot2" width="80%" />
<p class="caption">ggplot2</p>
</div>
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Scatter plots .tiny[(1/3)]

- Create simple scatter plot of bodymass vs. bill length

```r
plot(x = penguins$body_mass_g, 
     y = penguins$bill_length_mm, 
*    type = "p", cex = 1.0, pch = 19,
     xlab = "Body mass [g]", ylab = "Bill length [mm]")
```
]

.pull-right[
<img src="data:image/png;base64,#index_files/figure-html/scatter-1.png" width="80%" style="display: block; margin: auto;" />
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Scatter plots .tiny[(2/3)]

- Points can be colored based on values

```r
# specify color palette
color_palette <- c(Chinstrap = "#c15bcb", 
                   Gentoo = "#077476", 
                   Adelie = "#ff8200")

# make sure species factor has same order
penguins$species <- factor(penguins$species, 
                           levels = c("Chinstrap", 
                                      "Gentoo", 
                                      "Adelie"))

plot(x = penguins$body_mass_g, 
     y = penguins$bill_length_mm, 
*    col = color_palette[penguins$species],
     type = "p", cex = 1.0, pch = 19,
     xlab = "Body mass [g]", ylab = "Bill length [mm]")
```
]

.pull-right[
<img src="data:image/png;base64,#index_files/figure-html/scatter_color-1.png" width="80%" style="display: block; margin: auto;" />
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Scatter plots .tiny[(3/3)]

- Add other graphical objects to a plot

```r
# fit regression model (w/o species) 
lm_ttl <- lm(formula = bill_length_mm ~ body_mass_g, 
             data = penguins)

plot(x = penguins$body_mass_g, 
     y = penguins$bill_length_mm, 
     col = color_palette[penguins$species],
     type = "p", cex = 1.0, pch = 19,
     xlab = "Body mass [g]", ylab = "Bill length [mm]")

# add regression line
*abline(lm_ttl)

# add legend
*legend("topleft", legend = c("Chinstrap",
                             "Gentoo", 
                             "Adelie"), 
       col = color_palette, pch = 19)
```
]

.pull-right[
<img src="data:image/png;base64,#index_files/figure-html/scatter_lm-1.png" width="80%" style="display: block; margin: auto;" />
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Barplot vs. Boxplot

```r
df_sum <- dplyr::filter(penguins, species == "Adelie") %>% 
  dplyr::group_by(sex) %>% 
  dplyr::summarise(mn = mean(flipper_length_mm), 
                   sd = sd(flipper_length_mm))

*p_bar <- barplot(mn ~ sex, data = df_sum,
*       ylim = c(0, 220))

arrows(x0 = p_bar, y0 = df_sum$mn - df_sum$sd, 
       x1 = p_bar, y1 = df_sum$mn + df_sum$sd, 
       lwd = 1.5, angle = 90, code = 3)
```

<img src="data:image/png;base64,#index_files/figure-html/barplot-1.png" width="55%" style="display: block; margin: auto;" />
]

```r
boxplot(flipper_length_mm ~ sex, data = penguins, 
        subset = species == "Adelie")
```

<img src="data:image/png;base64,#index_files/figure-html/boxplot-1.png" width="65%" style="display: block; margin: auto;" />
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Histogram and Density .tiny[(1/3)]

```r
hist(penguins$bill_length_mm, breaks = 20)
```
]

.pull-right[
<img src="data:image/png;base64,#index_files/figure-html/hist-1.png" width="75%" style="display: block; margin: auto;" />
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Histogram and Density .tiny[(2/3)]

```r
chinstrap <- penguins$bill_length_mm[penguins$species == "Chinstrap"]
gentoo <- penguins$bill_length_mm[penguins$species == "Gentoo"]
adelie <- penguins$bill_length_mm[penguins$species == "Adelie"]

hist(chinstrap, breaks = 20, col = "#c15bcb", 
     xlim =c(30, 60), ylim = c(0, 20), 
     xlab = "Bill length [mm]", main = "")

hist(gentoo, breaks = 20, col = "#077476", 
*    add = TRUE)

hist(adelie, breaks = 20, col = "#ff8200", 
*    add = TRUE)
```
]

.pull-right[
<img src="data:image/png;base64,#index_files/figure-html/hist_species-1.png" width="75%" style="display: block; margin: auto;" />
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Histogram and Density .tiny[(3/3)]

```r
*hist(chinstrap, breaks = 20, probability = TRUE,
     border = "#c15bcb", col = NA, xlim = c(30, 60), 
     xlab = "Bill length [mm]", main = "")

hist(gentoo, breaks = 20, probability = TRUE, 
     border = "#077476", col = NA, add = TRUE)

hist(adelie, breaks = 20, probability = TRUE, 
     border = "#ff8200", col = NA, add = TRUE)

*lines(density(chinstrap), col = "#c15bcb")
*lines(density(gentoo), col = "#077476")
*lines(density(adelie), col = "#ff8200")
```
]

.pull-right[
<img src="data:image/png;base64,#index_files/figure-html/density-1.png" width="75%" style="display: block; margin: auto;" />
]

---

background-image: url("data:image/png;base64,#img/data-viz.png")
background-position: 50% 25%
background-size: 60%

## Exercise 3: Data vizualisation

... work on [_exercise-section-3.Rmd_](https://github.com/mhesselbarth/introduction-r-workshop/blob/main/exercises/exercise-section-3.Rmd) ...

---

## Section 4: Some more programming...

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Loops .tiny[(1/2)]

- Code is **repeated** until condition is met

- `for` and `while` loops most common

- Useful if vectorization is not possible, but **can be slow** if used incorrectly

- (Never grow a vector, always pre-allocate!)

```r
n <- 3
result_good <- numeric(length = n)
result_bad <- numeric()

for (i in 1:n) {
  
  x <- i * 5
  
* result_good[i] <- x
  result_bad <- c(result_bad, x)
  
  print(paste0("i=", i, " // x=", x))
}
## [1] "i=1 // x=5"
## [1] "i=2 // x=10"
## [1] "i=3 // x=15"
```
]

```loop_species
for (i in c("Chinstrap", "Gentoo", "Adelie")) {
  dplyr::filter(penguins, species == i) %>% 
    plot(body_mass_g ~ flipper_length_mm, 
         data = ., main = paste("Species:", i))
}
```
]

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# Loops .tiny[(2/2)]

- `apply()` (_matrices_) and `lapply()` (_vectors_, _data.frames_, _lists_) as alternative instead of loops

```r
penguins_mat <- as.matrix(penguins[, 3:6])

# calculate maximum of columns
*apply(penguins_mat, 2, max)
##    bill_length_mm     bill_depth_mm flipper_length_mm       body_mass_g 
##              59.6              21.5             231.0            6300.0

penguins_list <- dplyr::group_by(penguins, island) %>% 
  dplyr::group_split()

# use anonymous function to calculate cv of each island
lapply(penguins_list, function(i){ 
* data.frame(island = unique(i$island), cv = sd(i$body_mass_g) / mean(i$body_mass_g))})
## [[1]]
##   island        cv
## 1 Biscoe 0.1675845
## 
## [[2]]
##   island        cv
## 1  Dream 0.1110369
## 
## [[3]]
##      island        cv
## 1 Torgersen 0.1218404
```

---

background-image: url("data:image/png;base64,#img/r-logo.png")
background-position: 95% 5%
background-size: 10%

# `ifelse` statements

- General form: `if(conditions) {do this} else {do that}`

- Often used in _loops_ and _functions_

- Vectorized `ifelse()` function available
]

```r
# useing vectorized function
*ifelse(test = penguins$body_mass_g <= 3500,
*      yes = "small", no = "large") %>%
  head()
## [1] "large" "large" "small" "small" "large" "large"

# using loop and classic if else statement
class <- vector(mode = "character", length = nrow(penguins))

for (i in 1:nrow(penguins)) {
  
  # if body mass above 10000 something is probably wrong
  if (penguins[i, "body_mass_g"] > 10000) stop("Too big?")
  
  # classify based on threshold
* if (penguins[i, "body_mass_g"] <= 3500) {
    class[i] <- "small"
* } else {
     class[i] <- "large"
  }
}

head(class)
## [1] "large" "large" "small" "small" "large" "large"
```
]

---

# Resources

[_R_ for Data Science](https://r4ds.had.co.nz)

[Advanced _R_](https://adv-r.hadley.nz)

[Efficient _R_ programming](https://csgillespie.github.io/efficientR/)

[What They Forgot to Teach You About _R_](https://rstats.wtf/index.html)

[Big Book of _R_](https://www.bigbookofr.com/index.html)

[Modern Statistics for Modern Biology](https://www-huber.embl.de/msmb/index.html)

[Modern Statistics with R](http://www.modernstatisticswithr.com)

[Introduction to Modern Statistics](https://openintro-ims.netlify.app)

[ggplot2](https://ggplot2-book.org/index.html)

---

# References

.tiny[
Atkins, J.W., Stovall, A.E.L., Alberto Silva, C., 2022. _Open-Source tools in R for forestry and forest ecology._ Forest Ecology and Management 503, 119813. https://doi.org/10.1016/j.foreco.2021.119813

Esquivel, K.E., Hesselbarth, M.H.K., Allgeier, J.E., 2022. _Mechanistic support for increased primary production around artificial reefs._ Ecological Applications e2617. https://doi.org/10.1002/eap.2617

Hesselbarth, M.H.K., Sciaini, M., With, K.A., Wiegand, K., Nowosad, J., 2019. _landscapemetrics: An open‐source R tool to calculate landscape metrics._ Ecography 42, 1648–1657. https://doi.org/10.1111/ecog.04617

Hesselbarth, M.H.K., 2021. _shar: An R package to analyze species-habitat associations using point pattern analysis._ Journal of Open Source Software 6, 3811. https://doi.org/10.21105/joss.03811

Hesselbarth, M.H.K., Nowosad, J., Signer, J., Graham, L.J., 2021. _Open-source tools in R for landscape ecology._ Current Landscape Ecology Reports 6, 97–111. https://doi.org/10.1007/s40823-021-00067-y

Horst A.M., Hill A.P., Gorman K.B., 2020. _palmerpenguins: Palmer Archipelago (Antarctica) penguin data._ R package version 0.1.0.

Joo, R., Boone, M.E., Clay, T.A., Patrick, S.C., Clusella‐Trullas, S., Basille, M., 2020. _Navigating through the R packages for movement._ Journal of Animal Ecology 89, 248–267. https://doi.org/10.1111/1365-2656.13116

Lai, J., Lortie, C.J., Muenchen, R.A., Yang, J., Ma, K., 2019. _Evaluating the popularity of R in ecology._ Ecosphere 10. https://doi.org/10.1002/ecs2.2567
]

---

## Thank you for your attention.

### Questions?

<br>