intro-to-R.qmd

---
title: "Introduction to R"
subtitle: "<br>\nData Wrangling and Exploring Data with R"
author: "Data Academy"
execute:
  echo: true
format: 
  revealjs:
    code-block-background: true
    theme: [default, custom.scss]
    smaller: true
---

## Course Outline

:::: {.columns} 

::: {.column width="50%" .incremental}

Overview

* RStudio
* R Projects
* Variables
* Types and Data Structures
* Operators
* Functions
* Packages
* Reading/Writing Data
:::

::: {.column width="50%" .incremental}

Exploratory Data Analysis

* Data Frames
* The tidyverse
* Subsetting, Sorting, and Summarizing
* Manipulating
* Piping
* Joining
* Pivoting
* Plotting
* Script Structure

:::

::::

## Course Goals

:::: {.columns}

::: {.column width="40%" .incremental}

* Navigate RStudio
* Pledge to use R Projects
* Understand R's types and structures
* Manipulate data frames
* Conduct basic EDA
* Know where and how to get help

:::

::: {.column width="60%" height="75%" .incremental}

![](img/first_i_was_like.png){.absolute height="600"}

:::

::::


## Why use R?

:::: {.columns}

::: {.column width="40%" .incremental}

* It's powerful.
* It promotes reproducibility.
* It's popular around the city.
* It works well with other languages.
* There are packages for almost everything.
* If there isn't a package for something, it's fairly easy to make it yourself!
* It's cooler than Python.

:::

::: {.column width="60%" .incremental}

![](img/first_i_was_like.png){.absolute height="600"}

:::

:::

## Gallery



* "All Aboard! Exploring the Amtrak Passenger Rail System" [(table)](https://joshfangmeier.quarto.pub/table-contest-2022/#sec-table) [(code)](https://github.com/jfangmeier/table-contest-2022/blob/main/R/amtrak-table-standalone.qmd)

* "Locating neighborhood diversity in the American Metropolis" [(app)](https://walkerke.shinyapps.io/neighborhood_diversity/) [(code)](https://github.com/walkerke/neighborhood_diversity/blob/master/neighborhood_diversity.Rmd)

* [ggplot2](https://ggplot2.tidyverse.org/) and extensions

```{r echo=FALSE, eval=TRUE}
library(gapminder)
library(ggplot2)
library(gganimate)

ggplot(gapminder, aes(gdpPercap, lifeExp, size = pop, colour = country)) +
  geom_point(alpha = 0.7, show.legend = FALSE) +
  scale_colour_manual(values = country_colors) +
  scale_size(range = c(2, 12)) +
  scale_x_log10() +
  facet_wrap(~continent) +
  labs(title = 'Year: {frame_time}', x = 'GDP per capita', y = 'life expectancy') +
  transition_time(year) +
  ease_aes('linear')
```


## Your First R Project

(Live)

[Intro to R GitHub Repo](https://github.com/SFOEWD/data-academy-intro-to-r)

Use R Projects because:

* They promote sensible file/directory structure.
* They're easy to share.
* Manually fiddling with your working directory is bad practice.


## Variable Assignment

Variable names can't include spaces and must start with a letter. 

```{r}
a <- 5
a = 5
a
b <- 4
b
c <- a + b
c
test_scores <- c(82, 89, 92, 75, 74, 99)
test_scores

```

::: {.callout-tip}
`c()` is a function that 'concatenates' a vector. More on vectors in a bit!

:::

## Functions

Functions are pre-defined code that accomplish one specific task. A function has two components: (1) the name of the function; and (2) the input or 'arguments'. The value returned is called the 'output.' Running or executing a function is called 'calling' a function. 

```{r}
mean(x = c(3, 7, 11, 12, 14, 15))
```

* A function can have any number of arguments. Some are required, but some are optional. 
* If an argument is not explicitly specified, a default will be used.

```{r}
mean(c(3, 7, 11, 12, 14, 15, NA))
mean(c(3, 7, 11, 12, 14, 15, NA), na.rm = TRUE)
```

* A function will throw an error if a required argument is not used.

```{r error=TRUE}
mean(na.rm = TRUE)
```


## Types

:::: {.columns}

::: {.column width="25%"}

Types

* numbers
* characters
* logical
* dates
* factors

:::

::: {.column width="75%"}

:::

:::

## Types - numbers

:::: {.columns}

::: {.column width="25%"}

Types

* [numbers]{style="color:firebrick;"}
* characters
* logical
* dates
* factors

:::

::: {.column width="75%" .incremental}

R has three numeric types: numeric (or floating point), integer, and complex.

```{r}
n <- 2.5
class(n)

int <- as.integer(2)
class(int)

class(-1 + 3i)
```

:::

:::

## Types - characters

:::: {.columns}

::: {.column width="25%"}

Types

* numbers
* [characters]{style="color:firebrick;"}
* logical
* dates
* factors

:::

::: {.column width="75%" .incremental}

Characters are for storing text or strings and can be entered using either double or single quotes.

```{r error=TRUE}
greetings <- "hello world"

class(greetings)
```

'Escape' quotes within a character string with backslashes.

```{r}
my_enthusiasm <- 'I\'m loving R!'
print(my_enthusiasm)
```


:::

:::

## Types - boolean/logical

:::: {.columns}

::: {.column width="25%"}

Types

* numbers
* characters
* [logical]{style="color:firebrick;"}
* dates
* factors

:::

::: {.column width="75%" .incremental}

Logical types have three possible values: TRUE, FALSE, and NA.

```{r error=TRUE}
class(TRUE)
cond <- 1 > 2
cond
class(cond)
```

:::

:::

## Types - dates

:::: {.columns}

::: {.column width="25%"}

Types

* numbers
* characters
* logical
* [dates]{style="color:firebrick;"}
* factors

:::

::: {.column width="75%"}

Objects of the date class are used for dates.

```{r error=TRUE}
d <- as.Date("2024/09/07")
class(d)
as.Date("09/07/2024")
as.Date("09/07/2024", format = "%m/%d/%Y")
```

```{r echo=FALSE}
knitr::kable(data.frame(c("\\%Y", "\\%y", "\\%m", "\\%d"), c("4-digit year (1982)", "2-digit year (82)", "2-digit month (01)", "2-digit day of the month (13)")), col.names = NULL)
```

:::

:::

## Types - factors

:::: {.columns}

::: {.column width="25%"}

Types

* numbers
* characters
* logical
* dates
* [factors]{style="color:firebrick;"}

:::

::: {.column width="75%"}

Factors are used for categorical, 'ordinal' variables. 


```{r class}
m <- c("Dec", "Apr", "Jan", "Mar")
sort(m)

month_levels <- c(
  "Jan", "Feb", "Mar", "Apr", "May", "Jun", 
  "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"
)

m <- factor(m, levels = month_levels)
sort(m)
```

:::

:::

## Structures - vectors

:::: {.columns}

::: {.column width="25%"}

Types

* [vector]{style="color:firebrick;"}
* list
* matrix
* data frame

:::

::: {.column width="75%"}

A vector is a one dimensional collection of elements of the same class.  

```{r error=TRUE}
temps <- c(94, 83, 79, 55, 65)
temps

groceries <- c("apples", "carrots", "ice cream", "hot sauce")
groceries

conds <- c(TRUE, TRUE, FALSE, FALSE)
conds
```

::: {.callout-important}
Be wary of R's implicit coercion
:::

```{r}
c(1, 2, TRUE)
c("A", "B", TRUE)
```

:::

:::

## Subsetting vectors

:::: {.columns}

::: {.column width="25%"}

Types

* [vector]{style="color:firebrick;"}
* list
* matrix
* data frame

:::

::: {.column width="75%"}

Elements can be accessed using 'subscripts' or 'indices', which are specified using brackets:

```{r}
groceries <- c("apples", "carrots", "ice cream", "hot sauce")
groceries[2]
groceries[-2]
groceries[c(1, 2, 3)]
groceries[1:3]

```


:::

:::

## Structures - lists

:::: {.columns}

::: {.column width="25%"}

Types

* vector
* [list]{style="color:firebrick;"}
* matrix
* data frame

:::

::: {.column width="75%"}

A list can include elements of any type and of any dimension.

```{r error=TRUE}
l1 <- list("A", 1, TRUE)
l1
l2 <- list(c(1, 2, 3, 4, 5), c("a", "b", "c"), c(TRUE, TRUE))
l2
```


:::

:::

## Structures - matrices

:::: {.columns}

::: {.column width="25%"}

Types

* vector
* list
* [matrix]{style="color:firebrick;"}
* data frame

:::

::: {.column width="75%"}

A matrix is a two-dimensional collection of elements of the same type.

```{r error=TRUE}
m1 <- matrix(
  c(43, 43, 65, 76, 87, 34),
  nrow = 3,
  ncol = 2
)
m1

m2 <- matrix(
  c(TRUE, TRUE, FALSE, FALSE, FALSE, TRUE, TRUE, FALSE),
  nrow = 2,
  ncol = 4
)
m2
```


:::

:::

## Structures - data frames

:::: {.columns}

::: {.column width="25%"}

Types

* vector
* list
* matrix
* [data frame]{style="color:firebrick;"}

:::

::: {.column width="75%"}

A data frame is a table made of equal length vectors. 

```{r error=TRUE}
df <- data.frame(
  x = c(1, 2, 3),
  y = c(5, 6, 7),
  z = c("a", "b", "c")
)
df
```

* A 'tibble' is an improved type of data frame.

```{r}
library(tidyverse)
df <- tibble(
  x = c(1, 2, 3),
  z = c("a", "b", "c"),
  f = c(TRUE, FALSE, TRUE)
)
df
```


:::

:::

## Subsetting data frames

:::: {.columns}

::: {.column width="25%"}

Types

* vector
* list
* matrix
* [data frame]{style="color:firebrick;"}

:::

::: {.column width="75%"}

Use `$` to access columns (vectors) within a data frame.

```{r error=TRUE}
df$x
sum(df$x)
df$z[1]
df$z[c(1, 3)]
```

Subscripts/indices for data frames are pairs specifying the row and column numbers.

```{r}
df[1,1]
df[2, 1:3]
```


:::

:::

## Operators

```{r echo=FALSE}
gt::gt(tribble(
  ~Category, ~Operator,                               ~Operation, ~Example, 
  "Artithmetic",      "+",                          "Addition",  "x + y", 
  "Artithmetic",      "-",                       "Subtraction",  "x - y", 
  "Artithmetic",      "*",                    "Multiplication",  "x * y", 
  "Artithmetic",      "/",                          "Division",  "x / y", 
  "Artithmetic",      "^",                          "Exponent",  "x ^ y", 
  "Artithmetic",     "%%", "Modulus (Remainder from division)", "x %% y", 
  "Comparison",     "==",                    "Equal", "x == y",  
  "Comparison",     "!=",                "Not equal", "x != y",  
  "Comparison",      ">",             "Greater than",  "x > y",  
  "Comparison",      "<",                "Less than",  "x < y",  
  "Comparison",     ">=", "Greater than or equal to", "x >= y",  
  "Comparison",     "<=",    "Less than or equal to", "x <= y",
  "Logical",         "&",   "AND",                   "x & y",
  "Logical",         "|",   "OR",                   "x | y",
  "Logical",          "!",   "NOT",                 "!(x > y)",
  "Logical",         "%in%",  "IN",                  "x %in% y",
  "Sequence",        ":",     "Sequence",            "1:10"
  )) %>% gtExtras::gt_theme_538() %>% 
  gt::tab_options(table.background.color = "#d7e3f5")


```

## Vectorization

Many of R's operations are 'vectorized', meaning a given operation will operate on each element of a vector without explicit specification.

```{r}
x <- 1:5
x + 1
x < 4
x == 4

x[x < 4]

temps_f <- c(94, 83, 79, 71, 72)
temps_c <- (temps_f - 32)*5/9
temps_c
```


## Practice Using Functions!

```{r}
temps <- c(71, 91, 77, 81, 91, 91, 68, 66, 88, 82, 85, 76, 83, 91, 81, 77, 89, 68)
min(temps)
max(temps)
median(temps)
range(temps)
sd(temps)
sum(temps)
length(temps)
sum(temps > 90)
mean(temps > 90)
```

## Packages

Packages extend R's functionality beyond the functions available in the 'base' version. 

* Before you can use the functions from a package, the package must first be **installed.** 

* After a package has been installed, it can then be **loaded** into your session.

```{r eval=FALSE }
install.packages("dplyr") # you only need to install the package once
library(dplyr) # but you need to load the package each session 
```

The [Comprehensive R Archive Network](https://cran.r-project.org/) serves as a repository for most packages (21,1145 as of August 2024).

![](img/rpackages.png)

## The Tidyverse

"The tidyverse is an collection of R packages designed for data science. All packages share an underlying design philosophy, grammar, and data structures."

![](img/tidyverse.png)

## Key Tidyverse Packages

:::: {.columns}

::: {.column width="60%"}

* **readr** - functions for reading rectangular data (like csv, tsv, and fwf).
* **readxl** - functions to read data from .xls and .xlsx files.
* **dplyr** - a grammar of data manipulation that solve the most common data manipulation challenges.
* **tidyr** - tools for pivoting, cleaning, and 'tidying' data.
* **ggplot2** - a system for creating graphics, based on The Grammar of Graphics.
* **purrr** - tools for working with functions and vectors.
* **lubridate** - functions for working with dates.
* **stringr** - functions for working with strings.
* **forcats** - functions for working with factors.

:::

::: {.column width="40%"}

![](img/tidyverse-hex.png)

:::

:::

## Reading Data

```{r eval=FALSE}
library(readr)
penguins <- read_csv("data/penguins.csv")
```

```
Rows: 344 Columns: 8                                                                                                                                                     
── Column specification ──────────────────────────────────────────────────────────────────────────────────────────────────────────────
Delimiter: ","
chr (3): species, island, sex
dbl (5): bill_length_mm, bill_depth_mm, flipper_length_mm, body_mass_g, year

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
```


```{r eval=FALSE}
library(readxl)
penguins <- read_xlsx("data/penguins.xlsx")
penguins <- read_rds("data/penguins.rds")
```


## Reading from DataSF

* Source: [Traffic Crashes Resulting in Fatality](https://data.sfgov.org/Public-Safety/Traffic-Crashes-Resulting-in-Fatality/dau3-4s8f/about_data)

```{r}
library(RSocrata)
crashes <- read.socrata("https://data.sfgov.org/resource/dau3-4s8f.csv")
# crashes <- read_csv("https://data.sfgov.org/resource/dau3-4s8f.csv")
glimpse(crashes)


```

```{r echo=FALSE, message=FALSE}
library(tidyverse)
library(palmerpenguins)
options(pillar.print_max = 3)
options(pillar.print_min = 3)
```

## select()


```{r}
library(dplyr)
select(penguins, species, island, sex, body_mass_g)

select(penguins, 1:4)

select(penguins, bill_length_mm:body_mass_g)
```

## arrange()

```{r}
arrange(penguins, bill_length_mm)

arrange(penguins, species)

arrange(penguins, desc(bill_length_mm))
```

## rename()

```{r}
rename(penguins, Sex = sex)

rename(penguins, genus = species, isle = island)
```


## distinct()

```{r}
distinct(penguins, sex)

distinct(penguins, island)

distinct(penguins, island, species)
```


## filter()

```{r}
filter(penguins, sex == "female")

filter(penguins, body_mass_g > 4800)

filter(penguins, sex == "female", body_mass_g >= 4800)
```


## filter()

```{r}
filter(penguins, is.na(bill_length_mm))
filter(penguins, !is.na(bill_length_mm))
```

## filter()

```{r}
filter(penguins, island == "Biscoe" | island == "Dream")
filter(penguins, island %in% c("Biscoe", "Dream"))
filter(penguins, !island %in% c("Biscoe", "Dream"))
```

## filter() with string helpers

```{r}
library(stringr)

filter(penguins, str_detect(island, "ger"))

filter(penguins, str_length(species) == 6)
```


## mutate()

```{r}
mutate(penguins, body_mass_lb = body_mass_g/453.6)

usa_penguins <- mutate(
  penguins, 
  body_mass_lb = body_mass_g/453.6, 
  flipper_length_in = flipper_length_mm/25.4
  )

select(usa_penguins, species, body_mass_lb, flipper_length_in)
```

## mutate() with helpers

* If you need to specify only one condition, use `if_else()`. Otherwise, use `case_when()`.

```{r}
mutate(penguins, body_mass_g = if_else(island == "Biscoe", body_mass_g - 50, body_mass_g))

new_measurements <- mutate(penguins, new_body_mass_g = case_when(
  island == "Biscoe" ~ body_mass_g - 50,
  island == "Dream" ~ body_mass_g - 75,
  island == "Torgersen" ~ body_mass_g - 100
  )
)
select(new_measurements, island, body_mass_g, new_body_mass_g)
```

## mutate() with string helpers

```{r}
penguins_with_ids <- mutate(penguins, id = paste(island, species, sex, year, sep = "-"))
select(penguins_with_ids, island, species, sex, year, id)

mutate(penguins, sex = str_sub(sex, start = 1, end = 1))

mutate(penguins, sex = str_to_title(sex))
```


## count()

```{r echo=FALSE}
options(pillar.print_max = 8)
options(pillar.print_min = 8)
```


```{r}
count(penguins, sex)

count(penguins, species)

count(penguins, sex, species, sort = TRUE)
```

## count()

```{r}
count(penguins, island, name = "n_island_dwellers")

count(penguins, island == "Biscoe")

count(penguins, body_mass_g < 3000)
```


## summarize()

```{r}
summarize(penguins, mean_flipper_length = mean(flipper_length_mm))

summarize(
  penguins, 
  mean_flipper_length = mean(flipper_length_mm, na.rm = TRUE),
  mean_body_mass = mean(body_mass_g, na.rm = TRUE),
  mean_bill_length = mean(bill_length_mm, na.rm = TRUE),
  )
```

## group_by()

```{r}
penguins_grouped_by_sex <- group_by(penguins, sex)
summarize(penguins_grouped_by_sex, mean_body_mass = mean(body_mass_g, na.rm = TRUE))

penguins_grouped_by_sex_and_species <- group_by(penguins, sex, species)
summarize(penguins_grouped_by_sex_and_species, mean_body_mass = mean(body_mass_g, na.rm = TRUE))

summarize(penguins, mean_body_mass = mean(body_mass_g, na.rm = TRUE), .by = sex)
```

## The Pipe

We typically want to run numerous operations on a data frame, and saving the intermediate outputs as separate variables is tedious. The 'pipe' operator (`%>%` or `|>`), passes the output from one function directly into another. 

* Windows: `Ctrl + M`; Mac: `Cmd + M`

```{r}
penguins %>% 
  filter(body_mass_g > 4000) %>% 
  group_by(sex) %>% 
  summarize(mean_body_mass = mean(body_mass_g, na.rm = TRUE)) %>% 
  arrange(desc(mean_body_mass))
```

```{r}
penguins %>% count(species)
```

## The tidyverse paradigm

* 'siuba' in Python:
```{python eval=FALSE, python.reticulate = FALSE}
from siuba import group_by, summarize, _
from siuba.data import mtcars

(mtcars
  >> group_by(_.cyl)
  >> summarize(avg_hp = _.hp.mean()))
```

* 'Tidier.jl' in Julia:
```{julia eval=FALSE}
@chain movies begin
    @group_by(Year)
    @summarize(Mean_Yearly_Rating = mean(skipmissing(Rating)),
        Median_Yearly_Rating = median(skipmissing(Rating)))
    @slice(1:5)
end
```

* PRQL:
```{prql eval= FALSE}
from invoices
filter invoice_date >= @1970-01-16
derive {
  transaction_fees = 0.8,
  income = total - transaction_fees
}
```


## Practice using the tidyverse!

Source: [Air Traffic Passenger Statistics](https://data.sfgov.org/Transportation/Air-Traffic-Passenger-Statistics/rkru-6vcg/about_data)

```{r eval=FALSE}
air_traffic <- read.socrata("https://data.sfgov.org/resource/rkru-6vcg.csv")

# How many passengers deplaned from airlines with 'China' in their name?
air_traffic %>% 
  filter(
    str_detect(operating_airline, "China"),
    activity_type_code == "Deplaned"
    ) %>% 
  group_by(operating_airline) %>% 
  summarize(passengers = sum(passenger_count)) %>% 
  arrange(desc(passengers))

# How many flights for each operating airline in 2020?
air_traffic %>% 
  filter(
    activity_period_start_date >= as.Date("2020-01-01") & 
      activity_period_start_date <= as.Date("2020-12-31")
    ) %>% 
  count(operating_airline, sort = TRUE, name = "flights") %>% 
  head()
```

## left_join()

If a row in 'x' or the left-hand side matches a row in 'y' or the right-hand side, the columns from the y table are joined to the x table.

```{r}
df1 <- tibble(x = 1:3)
df2 <- tibble(x = c(1, 2), y = c("first", "second"))
df1
df2
left_join(df1, df2, by = "x") # or df1 %>% left_join(df2, join_by(x))
```


## left_join() (multiple matches)

If a row in 'x' or the left-hand side has **multiple** matches in 'y' or the right-hand side, **all** the matching rows in y will be joined to x. 


```{r}
df1 <- tibble(id = 1:3)
df2 <- tibble(code = c(1, 1, 2), y = c("first", "second", "third"))
df1
df2
df1 %>% left_join(df2, join_by(id == code))
```

## inner_join()

```{r}
x <- tibble(c1 = 1:3, c2 = c("x1", "x2", "x3"))
y <- tibble(c1 = c(1, 2, 4), c3 = c("y1", "y2", "y4"))
inner_join(x, y, by = join_by(c1))
```


![](img/inner-join.gif)


## bind_rows()

```{r}
penguins_2007 <- penguins %>% filter(year == 2007)
penguins_2008 <- penguins %>% filter(year == 2008)
nrow(penguins_2007)
nrow(penguins_2008)

all_penguins <- bind_rows(penguins_2007, penguins_2008)
all_penguins
```


## Practice joining data!

```{r eval=FALSE}
flights <- read_rds("data/flights.rds")
airlines <- read_rds("data/airlines.rds")
planes <- read_rds("data/planes.rds")
airports <- read_rds("data/airports.rds")

left_join(flights, airlines, by = join_by(carrier))

flights %>% 
  left_join(airports, join_by(dest == faa)) %>% 
  select(year, month, day, origin, dest, tzone)

flights %>% 
  inner_join(planes, join_by(tailnum)) %>% 
  select(flight, month, day, type, engine)
```

## pivoting

Reshape your data into something *longer* (increasing number of rows and decreasing the number of columns) or reshape your data into something *wider* (increasing the number of columns and decreasing the number of rows).

![](img/tidyr-pivoting.gif)


## pivot_longer()

```{r}
glimpse(relig_income)
relig_income %>% 
  pivot_longer(
    cols = 2:11,
    names_to = "income",
    values_to = "count"
  )
```


## pivot_wider()

```{r}
penguins %>% 
  count(island, species) %>% 
  pivot_wider(
    names_from = species,
    values_from = n
  )
```


## Writing Data

```{r eval=FALSE}
adelie_males_on_torgersen_in_2007 <- penguins %>% 
  filter(
    species == "Adelie",
    sex == "male",
    island == "Torgersen",
    year == "2007"
  ) %>% 
  select(bill_length_mm:body_mass_g)

write_csv(adelie_males_on_torgersen_in_2007, "data/adelie_males_on_torgersen_in_2007.csv")
write_rds(adelie_males_on_torgersen_in_2007, "data/adelie_males_on_torgersen_in_2007.rds")

library(writexl)
write_xlsx(adelie_males_on_torgersen_in_2007, "data/adelie_males_on_torgersen_in_2007.xlsx")

```

## ggplot2

```{r dev = "png", dev.args=list(bg="transparent")}
ggplot(data = penguins, aes(x = species, y = body_mass_g, fill = species)) +
  geom_boxplot(outliers = FALSE) +
  geom_jitter(width = 0.1, alpha = 0.3) +
  labs(x = "Species", y = "Body Mass (grams)", title = "Penguin Body Mass by Species") +
  theme_classic() +
  theme(legend.position = "none")
```


## gt

:::: {.columns}

::: {.column width="50%"}

```{r eval=FALSE}
library(gt)

penguins %>% 
  group_by(island, species, sex) %>% 
  summarize(
    mean_body_mass = mean(body_mass_g, na.rm = TRUE)
    ) %>% 
  ungroup() %>% 
  drop_na(sex) %>% 
  pivot_wider(
    names_from = sex, 
    values_from = mean_body_mass
    ) %>% 
  mutate(island = paste("On", island, "island")) %>% 
  rename(
    Island = island,
    Species = species,
    Female = female,
    Male = male
  ) %>% 
  gt(
    groupname_col = "Island", 
    rowname_col = "Species"
    ) %>% 
  tab_style(
    style = list(cell_text(align = "right")),
    locations = cells_stub(rows = TRUE)
  ) %>% 
  tab_header(
    title = "Penguin Body Mass",
    subtitle = "Adult penguins near Palmer Station"
  )
```


:::

::: {.column width="50%"}

```{r echo=FALSE}
library(gt)

penguins %>% 
  group_by(island, species, sex) %>% 
  summarize(mean_body_mass = mean(body_mass_g, na.rm = TRUE)) %>% 
  ungroup() %>% 
  drop_na(sex) %>% 
  pivot_wider(
    names_from = sex,
    values_from = mean_body_mass
  ) %>% 
  mutate(island = paste("On", island, "island")) %>% 
  rename(
    Island = island,
    Species = species,
    Female = female,
    Male = male
  ) %>% 
  gt(groupname_col = "Island", rowname_col = "Species") %>% 
  tab_style(
    style = list(
      cell_text(align = "right")
    ),
    locations = cells_stub(rows = TRUE)
  ) %>% 
  tab_header(
    title = "Penguin Body Mass",
    subtitle = "Adult foraging penguins near Palmer Station"
  ) %>% 
  tab_options(table.background.color = "#d7e3f5")
```

:::

:::



## Script Structure and Basic EDA

[eda.R]()


1. Load packages
2. Import raw data
3. Clean/reshape/'tidy' data
4. Analyze
5. Write cleaned data

## Where to get Help


:::: {.columns}

::: {.column width="30%"}

* Stackoverflow

* Posit Community

* Twitter/X/Mastadon/BlueSky

* CCSF Teams Channel

:::

::: {.column width="70%"}

![](img/debugging.png)

:::

:::



## How to get Help

:::: {.columns}

::: {.column width="30%"}

* [Reprex Do's and Don'ts](https://reprex.tidyverse.org/articles/reprex-dos-and-donts.html)

:::

::: {.column width="70%"}

![](img/reprex.png)

:::

:::


## What should I learn next?

* dplyr selection helpers:
  * `starts_with()`/`ends_with()`
  * `contains()`/`matches()`
  * `first_col()`/`last_col()`
  * `everything()`
  * `across()`
  * `where()`

* Reports and dashboards: [Tutorial: Hello, Quarto](https://quarto.org/docs/get-started/hello/rstudio.html)

* Everything about ggplot2: [ggplot2: Elegant Graphics for Data Analysis](https://ggplot2-book.org/)

* Everything about gt: [Introduction to Creating gt tables](https://gt.rstudio.com/articles/gt.html)

* Writing good functions: [Chapter 19, 'Functions', in *R for Data Science*](https://r4ds.had.co.nz/functions.html)

* Working with databases: [Chapter 21, 'Databases', in *R for Data Science*](https://r4ds.hadley.nz/databases)

* Spatial Stuff: [*Geocomputation in R*](https://r.geocompx.org/intro)

* Watch and learn from a pro: [David Robinson's Tidy Tuesday screencasts](https://www.youtube.com/playlist?list=PL19ev-r1GBwkuyiwnxoHTRC8TTqP8OEi8)

## Other Cool R Things

* Interactive JavaScript visualizations in R: [htmlwidgets gallery](https://gallery.htmlwidgets.org/) 

* Interactive web applications: [Welcome to Shiny](https://shiny.posit.co/r/getstarted/shiny-basics/lesson1/index.html)

* Package development: [R packages](https://r-pkgs.org/)

* Automating data pipelines: [The targets user manual](https://books.ropensci.org/targets/) 

* Make any chart: [The R Graph Gallery](https://r-graph-gallery.com/)

## Lab

[Lab instructions in the README here.](https://github.com/SFOEWD/data-academy-intro-to-r?tab=readme-ov-file#lab)

## fin

Give us your feedback! (Please respond to the survey sent out after class)

[](img/r-flowers.png)