How to Work with Data in R

This article describes how to access your data using R code and use it in calculations or other manipulations. This includes:

Accessing your data

Working with different data structures

1. Vector = a single row/column or a series of values/strings (most commonly a single variable or single column table in Displayr)

2. Matrix = a variable set with more than one variable and a table showing 1 statistic of one type (either  all cells are numeric or character)

3. Array = a table showing more than 1 statistic (more than 1 number in each cell)

4. Data.frame = a table showing 1 value in each cell that can have different types of data in the cells

Requirements

• Some of the functions below may require a Displayr license.
• Familiarity with the different Structures and Value Attributes for Variable Sets
• A Calculation or R variable
• Knowledge of How to Use Different Types of Data in R

Method - Accessing your data

There are several ways to access your data within R code. For more detail on our point-and-click functionality see How to Use Point and Click Inside R Code.

1. By dragging and dropping the variables from the Data Sources or Reports tree:
   
2. By clicking on rows/columns in a table:
   
   
3. By the Label inside backticks:
   
   
4. By the Name of the variable (found by hovering over the variable in the Data Sources tree as in the screenshot above) or an output on your page (found in General > Name):
   

Though you cannot reference a Data Set as a whole in the R code, if you have a variable that is named the same in multiple Data Sets, you will need to specify which Data Set to get it from by adding a prefix:

When a variable is referred to in R Code, the variable is automatically uploaded to the R Server prior to any R Code being run. A variable will have the following attributes:

• name. This is the Variable Name.
• question. This is the Question Name.
• label. This is the Variable Label.

While these attributes can be accessed in R in the usual way (e.g., attr(my.variable, "label"), the best way to access them is often using flipFormat::Labels, which will attempt to construct a label of form Question Name: Variable Label where these are different, and Variable Label where these two are the same (e.g., flipFormat::Labels(Q3) will show Q3. Age). It falls back to name, and, if even this is not provided, it attempts to discern the original name of the argument.

Method - Working with different data structures

You can reference a particular bit of your data structure using square brackets [ ] and the appropriate index.

1. Vector

A vector is a series of data points that can be any one data type (character, numeric, etc), but not be a mix of types (otherwise they will convert everything to character). When you reference a single variable, it will be in the form of a vector. One-column tables are also interpreted by R as vectors.

Vectors can also be created manually using the c() function:

numbers = c(2,5,10)

strings = c("hello", "good day", "good bye")

For numeric vectors, you can also create vectors with sequential numbers using :

#numbers 1 through 10
fingers = 1:10

Or numbers at set intervals using seq():

#even numbers 1 through 10
evens = seq(from=2, to=10, by=2)

For strings, those are usually automatically generated using paste or by pulling off the row or column headers from a table, see How to Use Paste Functions to Create Dynamic Text Using R.

Detailed example:

In the below example, we have a table called fruit:

Referencing:

• The syntax for indexing is: fruit[Item] .
• To return the value for Pear, we can use the row number fruit[2] or the row label fruit["Pear"].
• To fill in missing data for values under 5, we can use a condition inside brackets 
  fruit[fruit < 5] = NA.
  

Other Useful Functions:

• To return the row labels, we use names(fruit).
• To return the number of rows, we use length(fruit) or NROW(fruit).

2. Matrix

A matrix is a table with rows and columns where data is the same data type. In Displayr, built-in tables showing a single statistic, variable sets, and those created by cbind in R will be interpreted as matrices.

This can be created manually using the matrix() function:

tab = matrix(c(c(1,2,3), c("a","b","c")), ncol=2, nrow=3)

Detailed example:

In the below example, we have a crosstab table called living.alone:

Referencing:

• The syntax for indexing is:
  • living.alone[Row] for a single-column SUMMARY table.
  • living.alone[Row , Column] for any other table.
• To return the value for Male, we can use living.alone[1,] or living.alone["Male",].
• If there is only one column, use living.alone[Row , Column , drop = F] to keep the original table dimensions. Otherwise, the result will be interpreted as a vector.

Other Useful Functions:

• To return the row labels, we use rownames(living.alone).
• To return the column labels, we use colnames(living.alone).
• To return the table's dimensions we usedim(living.alone). This will return 3 (rows) and 1 (column).
• To return the number of rows, we useNROW(living.alone).
• To return the number of columns, we useNCOL(living.alone).

3. Array

An array is a multi-layered table where data is the same data type. In Displayr this is a crosstab with multiple statistics.

This can be created manually using the array() function:

tab = array(c(1,2,3), dim=c(3,4,2))

Detailed example:

In the below example, we have a table called living.alone with two statistics:

Referencing

• The syntax for indexing is: living.alone[Row , Column , Statistic] .
• To return the Count value for Male, we can use living.alone[1,,2] or living.alone["Male",,"Count"].
• If there is only one column, use living.alone[Row, Column, Statistic, drop = F] to keep the original table dimensions. 

Other Useful Functions:

• To return the table's dimensions we usedim(living.alone). This will return 3 (rows), 1 (column), and 2 (statistics).
• To return the row labels, we use rownames(living.alone) or
  dimnames(living.alone)[[1]].
• To return the column labels, we use colnames(living.alone)or
  dimnames(living.alone)[[2]].
• To return the statistic labels, we usedimnames(living.alone)[[3]].
• To return the number of rows, we useNROW(living.alone).
• To return the number of columns, we useNCOL(living.alone).

4. Data.frame

A data.frame is a table with rows and columns, like a matrix, but can be a mix of different types of data.

This can be created manually using the data.frame() function:

mydf = data.frame(Numbers=c(1,2,3), Letters=c("a","b","c"))

Detailed example:

Referencing and other useful functions are the same as used when working with a matrix, with some additional functionality below.

You can additionally reference an entire column using $. For example, mydf$Letters would return only the Letters column.

You can also add new columns on the fly with $:

mydf$`More Letters`=c("d","e","f")
mydf

Get a copy of the examples above in your account by clicking HERE.

See Also

How to Learn R

Using R in Displayr Video Series

How to Use R in Displayr

How R Works Differently in Displayr Compared to Other Programs

How to Use Different Types of Data in R

How to Use Point and Click Inside R Code

How to Reference Different Items in Your Document in R

How to Work with Conditional R Formulas

How to Extract Data from a Single Column Summary Table

How to Extract Data from a Multiple Column Table

How to Extract Data from a Multiple Column Table with Multiple Statistics

How to Extract Data from a Multiple Column Table with Nested Data