• So far we have entered everything directly in the Console
  \(\implies\) This is a bad idea in the real world
  • We have no record of what we've done
  • How do we remember anything 6 months later?
  • How do we submit our methods to our patent lawyer?
  • How do you explain your analysis to a collaborator
• We usually record everything as a script
  • Not just best practice \(\implies\) essential practice

• File > New File > R Script
  • A new text file will now be in the top left quarter
  • Save this as loadingData.R
• Giving it the suffix .R:
  • Helps RStudio with auto completion
  • Helps us find our R scripts in a messy folder
• We enter text and R commands in this file
  • They are not executed until we send them to the Console

• No formal requirements (unlike bash)
• Usually have a mix of R commands, and comments
  • Comments explain what you're intending to 'future' you and others
  • Comments in an R script start with #
  • Anything following # are ignored by the Console
• I usually start mine by loading packages

• Similar functions are collected into packages (e.g. base)
  • R always loads stats, graphics, grDevices, utils, datasets, methods and base at start-up
  • NEVER use the Packages Tab to load packages
  • Remember, we want a record of literally everything we do
    • Tools > Global Options > Packages > Uncheck Enable Packages Pane
• Let's start by loading some packages
  • Lloyd will take you through the dplyr package later
  • Contains functions for working with the data we're about to load

library(dplyr)

NB: We have not yet sent this command to the Console

1. Go to this line in the script
2. Linux/Windows: Ctrl + Enter or Mac Command + Enter
   • Ignore the friendly and helpful message
3. We have now loaded our first package…

NB: If you don't have dplyr installed:
Tools > Install Packages... > then type dplyr and hit Install

RStudio has a handy GUI for loading data into your R Environment

1. Go the the Environment Tab
2. Import Dataset ()
3. From CSV...
4. Browse to data/toothData.csv

Stop there!!!

1. Select all the code in the Code Preview Box
2. Paste this into your script, below library(dplyr)
3. Click Import
4. Magic happens…

The code we copied has 3 lines:

1. library(readr)
   • This package contains the function read_csv()
2. toothData <- read_csv("~/Documents/R-Training-20170501/data/toothData.csv")
   • This is the code executed to load the data into your R Environment
   • It created an object named toothData by using the file name.

1. library(readr)
   • This package contains the function read_csv()
2. toothData <- read_csv("~/Documents/R-Training-20170501/data/toothData.csv")
   • This is the code executed to load the data into your R Environment
   • It created an object named toothData by using the file name.
3. View(toothData) Open a preview in a familiar Excel-like format
   • I never use this
   • Close the preview by clicking the cross and delete View(toothData)

toothData

## # A tibble: 60 × 3
##      len  supp  dose
##    <dbl> <chr> <chr>
## 1    4.2    VC   Low
## 2   11.5    VC   Low
## 3    7.3    VC   Low
## 4    5.8    VC   Low
## 5    6.4    VC   Low
## 6   10.0    VC   Low
## 7   11.2    VC   Low
## 8   11.2    VC   Low
## 9    5.2    VC   Low
## 10   7.0    VC   Low
## # ... with 50 more rows

• A tibble is a type of R object called a data.frame
  • Used to be known as a tbl_df based on an SQL table
  • Became shortened to tibble
• Is a normal data.frame with a couple of nicer feature
  • Only prints what fits on your screen
  • A regular data.frame dumps everything

• These are very much like an Excel spreadsheet
  • Here we have 60 rows and 3 columns
• Each column is a vector

Q: What does this last statement mean?

• These are very much like an Excel spreadsheet
  • Here we have 60 rows and 3 columns
• Each column is a vector

Q: What does this last statement mean?

A : Now we have to care about what columns contain numbers, words etc.!

• We can just grab a single column by placing a $ after the name

toothData$len
toothData$supp
toothData$dose

• Lloyd will take us through subsetting in more detail

• Here supp and dose represent the supplement method and dosage of vitamin C given to guinea pigs
  • len represents the length of their incisors
• Would supp and dose be categorical variables?
  • Yes, let's convert them
  • Enter these lines in your script, then send to the Console

# Convert the supplement method and dosage to factors
toothData$supp <- as.factor(toothData$supp)
toothData$dose <- as.factor(toothData$dose)

toothData

• Note that the columns are now specified as <fctr>
• The old import method used a function read.csv()
  • Automatically converted characters to factors
  • Nearly everyone has made mistakes as a result
  • Best practice: If you want a factor \(\implies\) manually specify this!

Q: What if we changed our mind and didn't want them to be converted to factors?

Q: What if we changed our mind and didn't want them to be converted to factors?

A: Delete those lines from the script, and run it from the top again.

• Always copy those lines from the import GUI
• See how useful a script might be…

A data.frame is a special case of an R object called a list

• Beyond the scope of today

• In a data.frame, each column was a vector
  • Each column could be a different type of value
• A matrix is like a 2D vector
  • We have rows and columns
  • The entire matrix must be the same type of value
  • Can be very useful for statisticians and mathematicians
• Importantly, we cannot select a column of a matrix using the $ symbol

Q: What would happen if we converted toothData into a matrix?

as.matrix(toothData)

Q: What would happen if we converted toothData into a matrix?

A: The numeric values would become character values

• Vectors (1D) and Matrices (2D) are all the same type of values
• Data Frames (2D) have columns as vectors
• Always keep a script
  • Start a script with loading your packages
  • Write comments to communicate your intentions, and explain details
• Package installation uses install.packages("packageName")