Good morning!

We are Melinda, Vallen, Jaime, Kristie & Dorian.

We like R.

We aren’t computer scientists and that’s okay!

We make lots of mistakes. Mistakes are funny. You can laugh with us.

Create and store values

You can assign values to new objects using the “left arrow”, which is written as <-.

Left arrow

x <- 5

This is typed using a less-than sign followed by a hyphen. It’s more officially known as the assignment operator. Try adding the code below to your R script and assign a value to an object called partner.

my_partner <- "Partner's Name" # Text and characters are put in quotes

miles_traveled <- 1160 # A number has no quotes

# A list of data I use
data_types <- c("PAHs", "Ozone", "Fine particles") 

best_snack_ever_99 <- "Air Heads" 

my_partner

## [1] "Partner's Name"

nickname <- my_partner

nickname

It’s ALL about you

Now we can create a data table, which in R is also called a data.frame or a tibble. When creating one, the column names go on the left, and the values you want to put in the column go on the right.

# Put the items into a table
all_about_you <- data.frame(name           = my_partner, 
                            miles_traveled = miles_traveled, 
                            data_types     = data_types,  
                            best_snack     = best_snack_ever_99)

Let’s bounce around the room and introduce ourselves with help from our new data frames.

GET R PACKAGES

To use a new package in R you first need to install it – much like a free App on your phone. To save time on installation, you can copy the text below and paste it into the RStudio Console. It’s the quadrant on the lower left when you open RStudio. The one with the > symbols.

new_packages <- c("readr", "readxl", "dplyr", "stringr",
                  "ggplot2", "lubridate", "janitor", "curl")

install.packages(new_packages)

Then press ENTER to begin the installation. If all goes well, you should start to see some messages appear similar to these:

Congrats rebel droid!

R Community

• See the R Community page.
• ITEP page for sharing & questions - R questions
• Finding R Help - Get help!
  • R cheatsheets

When do we use R?

• To connect to databases
• To read data from websites
• To document and share methods
• When data will have frequent updates
• When we want to improve a process over time

R is for reading

Lucky for us, programming doesn’t have to be a bunch of math equations. R allows you to write your data analysis in a step-by-step fashion, much like creating a recipe for cookies. And just like a recipe, we can start at the top and read our way down to the bottom.

It begins!

Today’s challenge

Rey needs to visit Tatooine to help the Rebel Alliance, and we have ozone data to help her decide what month she should visit. Preferably the month with the lowest ozone concentrations. Let’s make a nice reference chart of monthly ozone concentrations to help her plan.

We’ll follow the general roadmap below.

Clean header names

There are two great packages that can help us with cleaning header names. Let’s install them!

install.packages("janitor")
install.packages("dplyr")

library(janitor)
library(dplyr)

# General cleaning for all columns
air_data <- clean_names(air_data)

# Change and set specific names
air_data <- rename(air_data,
                   lat = latitude,
                   lon = longitude)

names(air_data)

## [1] "date_time" "site"      "ozone"     "lat"       "lon"       "temp_f"   
## [7] "units"

Break it down now

glimpse() and summary()

Use the glimpse() function to find out what type and how much data you have.

library(dplyr)

# Glimpse the columns of your data and their first few contents
glimpse(air_data)

## Observations: 6,665
## Variables: 7
## $ date_time <dttm> 2015-07-06 09:00:00, 2015-07-13 04:00:00, 2015-07-1...
## $ site      <chr> "27-017-7417", "27-017-7417", "27-017-7417", "27-017...
## $ ozone     <dbl> 0, -1, -2, 0, 0, 0, 0, 0, 0, 0, -1, 0, -1, -1, -1, 0...
## $ lat       <dbl> 46.71369, 46.71369, 46.71369, 46.71369, 46.71369, 46...
## $ lon       <dbl> -92.51172, -92.51172, -92.51172, -92.51172, -92.5117...
## $ temp_f    <dbl> 64.36, 64.49, 63.23, 61.49, 61.30, 71.22, 76.63, 64....
## $ units     <chr> "PPM", "PPM", "PPM", "PPM", "PPM", "PPM", "PPM", "PP...

Use the summary() function to get a quick report of your numeric data.

# Show numeric summary of the min, mean, and max of all columns
summary(air_data)

##    date_time                       site               ozone         
##  Min.   :2014-04-03 03:00:00   Length:6665        Min.   :-4.00000  
##  1st Qu.:2015-06-07 10:00:00   Class :character   1st Qu.: 0.01700  
##  Median :2015-08-16 20:00:00   Mode  :character   Median : 0.02500  
##  Mean   :2015-08-13 15:04:35                      Mean   : 0.01666  
##  3rd Qu.:2015-10-24 12:00:00                      3rd Qu.: 0.03500  
##  Max.   :2016-01-01 05:00:00                      Max.   : 0.07500  
##       lat             lon             temp_f         units          
##  Min.   :46.71   Min.   :-92.54   Min.   : 1.83   Length:6665       
##  1st Qu.:46.71   1st Qu.:-92.51   1st Qu.:37.85   Class :character  
##  Median :46.71   Median :-92.51   Median :52.32   Mode  :character  
##  Mean   :46.73   Mean   :-92.51   Mean   :50.85                     
##  3rd Qu.:46.71   3rd Qu.:-92.51   3rd Qu.:63.59                     
##  Max.   :47.52   Max.   :-92.51   Max.   :90.01

Try running some of these in your script.

nrow(air_data)

ncol()

names()

It’s time for dplyr

This is our go-to package for most analysis tasks. With the six functions below you can accomplish just about anything you want.

Your new analysis toolbox

Function	Job
select()	Select individual columns to drop or keep
arrange()	Sort a table top-to-bottom based on the values of a column
filter()	Keep only a subset of rows depending on the values of a column
mutate()	Add new columns or update existing columns
summarize()	Calculate a single summary for an entire table
group_by()	Sort data into groups based on the values of a column

Convert units

mutate()

For mutate, the name of the column goes on the left, and the calculation of its new value goes on the right.

Update the column ozone

# Convert all samples to PPB
air_data <- mutate(air_data, ozone = ozone * 1000) 

Update the column units

# Set units column to PPB
air_data <- mutate(air_data, units = "PPB") 

---

The lubridate package

It’s about time! Lubridate makes working with dates easier. We can find how much time has elapsed, add or subtract days, and aggregate to seasonal, monthly, or day of the week averages.

Clean the dates

Let’s set our date column to the standard date format. Because our dates are written as year-month-day hour:mins, we can Use ymd_hm().

Format the date_time column as a ‘Date’

library(lubridate) 

# Set date column to official date format
air_data <- mutate(air_data, date_time = ymd_hms(date_time))

Now we can add a variety of date and time columns to our data like the name of the month, the day of the week, or just the hour of the day for each observation.

Add month and day of the week

# Add date parts as new columns
air_data <- mutate(air_data, 
                   year     = year(date_time),
                   month    = month(date_time, label = TRUE),
                   day      = wday(date_time, label = TRUE),
                   hour     = hour(date_time),
                   cal_date = date(date_time)) 

Polar plots

When looking at air concentration data, we often want to know what direction the wind is blowing from when an air pollutant tends to be elevated. This can help to answer if the pollution source is local or if it is more of a regionl issue.

Pairing wind direction and air concentration data helps answer these questions and provide further insights. Polar plots are one way to look at wind data and get to know the wind patterns around your air monitors.

Let’s make sure we’re looking at the data for only one of the sites.

ozone_met <- filter(ozone_met, site == "27-017-7417")

ggplot(ozone_met, aes(x = wd, y = ozone)) + 
      coord_polar() +
      geom_point(size = 2, alpha = 0.3) +
      scale_x_continuous(breaks = seq(0, 360, by = 90), 
                         lim    = c(0, 360), 
                         label  = c("","E", "S", "W", "N")) +
      theme_minimal()

ggplot(ozone_met, aes(x = wd, y = ozone, color = ozone)) + 
      coord_polar() +
      geom_point(size = 2, alpha = 0.3) +
      scale_x_continuous(breaks = seq(0, 360, by = 90), 
                         lim    = c(0, 360),
                         label  = c("","E", "S", "W", "N")) + 
      theme_minimal() + 
      scale_color_viridis_c()

Calendar plot

Sometimes with air data we want to know if there is seasonality in the data, or if there were dates with very high values. Calendar plots are great for this.

Let’s make this plot for only the year 2015. To do that we’ll first use filter().

ozone_met <- filter(ozone_met, year == 2015)

ozone_met <- filter(ozone_met, cal_date < "2016-01-01", cal_date > "2014-01-01")

ggplot(ozone_met, aes(x = day, y = hour, fill = ozone)) +
  geom_tile(color = "gray") 

ggplot(ozone_met, aes(day, month, fill = ozone)) +
  geom_tile(color = "gray")

ggplot(ozone_met, aes(hour, day, fill = ozone)) +
  geom_tile(color = "gray") +
  facet_wrap(~month)

miss_data <- filter(ozone_met,
                          day  ==  ______,
                          hour >   ______,
                          hour <   ______ )

Share on GitHub

Congratulations!

You’ve added some great tools to your data analysis tool belt. Now go forth and put them to use.

More data awaits you…

Make it your own

Let’s add a little style so R Studio feels like home since you will spend lots of time here. Follow these steps to change the font-size and and color scheme:

1. Go to Tools on the top navigation bar.
2. Choose Global Options...
3. Choose Appearance with the paint bucket.
4. Find something you like.