• A stumbling block for many learning R is the Error Messages
• We often see them while we're loading data
• R is very strict about data formats
• We can load .xlsx, xls, csv, txt, gtf/gff files + many more
• The structure of the spreadsheet is vital

• Things we do to make it "look nice" in Excel can create problems
• In Day1/data, open RealTimeData.xlsx in Excel (or Libre Office)

Which sheet do you think will be the most problematic to load?

Sheet 1

• This is actually the type of format R loves to see
• Simple column structure, with column and row names
• No blank rows at the top or separating sub-tables
• No blank columns

Sheet 2

What about all those missing values?

Sheet 2

What about all those missing values?

• R can happily deal with missing values: \(\implies\) will load as NA
• The missing column names may give strange behaviours
• Otherwise no problems

Sheet 3

• Here we effectively have 2 tables on the same sheet
  • Will this cause error messages?
  • R guesses the number of columns from the first row
  • The plot will simply be ignored

Always think in terms of columns

• In today's data folder is the file toothData.csv
• Clicking on it will open it in the Script Window as a text file
• We can use the button as shown

1. Try changing a few settings & note changes in the Preview
2. Once you're happy, click "Import"

You will see two lines of code in the Console \(\implies\) two things have just happened

toothData <- read.csv("RAdelaideWorkshop/Day_1/data/toothData.csv")
View(toothData)

toothData <- read.csv("RAdelaideWorkshop/Day_1/data/toothData.csv")

ALWAYS copy the first line into your script!

• This is the exact command we've used to load the file
• Anything we've set by clicking will also be in this line

View(toothData)

The second line has opened a preview of our R object

• By default, the R object will be named using the file-name before the .csv
• Also look at the toothData object in the Environment tab (click the arrow)

• The object toothData is known as a data.frame
• R equivalent to a spreadsheet
• Three possible ways to inspect this are:

View(toothData)
toothData
head(toothData)

What were the differences between each method?

• By default R assumes that a column of text is a categorical variable (i.e. a factor)
• We can change this by un-checking the stringsAsFactors button during import

• In the above we called the R function read.csv()
• From the utils package which is one of the base packages
• The help page isn't very helpful here…

?read.csv

• There are three basic functions listed here:
  read.table(), read.csv() and read.delim()
• The main differences are in the separators

• The above uses the default function read.csv()
• The package readr has a similar, but slightly superior version called read_csv()
• No GUI for this method

library(dplyr)
library(readr)
toothData <- read_csv("data/toothData.csv")

1. Faster
2. Character columns are left as plain text
3. It also runs some checks on your data \(\implies\) can give more error messages
4. Output is a local data frame \(\implies\) display in the Console is more convenient

toothData

?read_csv

• Once again we have four functions but stick to read_csv()
• This function has numerous arguments given as names
  (e.g. file, col_names)
  • If just the name is given, we need to specify something
  • Otherwise the argument is given a default value
    (e.g. col_names = TRUE)
  • If specifying in order, we don't need to name arguments

toothData <- read_csv("data/toothData.csv")

Is equivalent to:

toothData <- read_csv(file = "data/toothData.csv")

If we had a file with 3 blank lines to start, what would our code look like?

toothData <- read_csv("data/toothData.csv", ???)

If we had a file with 3 blank lines to start, what would our code look like?

toothData <- read_csv("data/toothData.csv", skip = 3)

What if the first three lines were comments starting with #?

If we had a file with 3 blank lines to start, what would our code look like?

toothData <- read_csv("data/toothData.csv", skip = 3)

What if the first three lines were comments starting with #?

• It could be the same.
• R uses the first row to determine the number of columns

toothData <- read_csv("data/toothData.csv", comment = "#")

• The bottom three functions are simplified wrappers to read_delim()
• read_csv() calls read_delim() using delim = ","
• read_csv2() calls read_delim() using delim = ";"
• read_tsv() calls read_delim() using delim = "\t"

What function would we call for space-delimited files?

R also has a package for loading .xls and xlsx files.

library(readxl)

The main function is read_excel()

?read_excel

Try loading each of the sheets from RealTimeData.xslx

(Remember to call the R objects something)

Do you get any error messages for sheet 3?

How could we load these two separate tables?