The easiest way to create summary tables in R is to use the describe() and describeBy() functions from the psych library.
library(psych) #create summary table describe(df) #create summary table, grouped by a specific variable describeBy(df, group=df$var_name)
The following examples show how to use these functions in practice.
Example 1: Create Basic Summary Table
Suppose we have the following data frame in R:
#create data frame df frame(team=c('A', 'A', 'B', 'B', 'C', 'C', 'C'), points=c(15, 22, 29, 41, 30, 11, 19), rebounds=c(7, 8, 6, 6, 7, 9, 13), steals=c(1, 1, 2, 3, 5, 7, 5)) #view data frame df team points rebounds steals 1 A 15 7 1 2 A 22 8 1 3 B 29 6 2 4 B 41 6 3 5 C 30 7 5 6 C 11 9 7 7 C 19 13 5
We can use the describe() function to create a summary table for each variable in the data frame:
library(psych) #create summary table describe(df) vars n mean sd median trimmed mad min max range skew kurtosis team* 1 7 2.14 0.90 2 2.14 1.48 1 3 2 -0.22 -1.90 points 2 7 23.86 10.24 22 23.86 10.38 11 41 30 0.33 -1.41 rebounds 3 7 8.00 2.45 7 8.00 1.48 6 13 7 1.05 -0.38 steals 4 7 3.43 2.30 3 3.43 2.97 1 7 6 0.25 -1.73 se team* 0.34 points 3.87 rebounds 0.93 steals 0.87
Here’s how to interpret each value in the output:
- vars: column number
- n: Number of valid cases
- mean: The mean value
- median: The median value
- trimmed: The trimmed mean (default trims 10% of observations from each end)
- mad: The median absolute deviation (from the median)
- min: The minimum value
- max: The maximum value
- range: The range of values (max – min)
- skew: The skewness
- kurtosis: The kurtosis
- se: The standard error
It’s important to note that any variable with an asterisk (*) symbol next to it is a categorical or logical variable that has been converted to a numerical variable with values that represent the numerical ordering of the values.
In our example, the variable ‘team’ has been converted to a numerical variable so we shouldn’t interpret the summary statistics for it literally.
Also note that you can use the argument fast=TRUE to only calculate the most common summary statistics:
#create smaller summary table describe(df, fast=TRUE) vars n mean sd min max range se team 1 7 NaN NA Inf -Inf -Inf NA points 2 7 23.86 10.24 11 41 30 3.87 rebounds 3 7 8.00 2.45 6 13 7 0.93 steals 4 7 3.43 2.30 1 7 6 0.87
We can also choose to only compute the summary statistics for certain variables in the data frame:
#create summary table for just 'points' and 'rebounds' columns describe(df[ , c('points', 'rebounds')], fast=TRUE) vars n mean sd min max range se points 1 7 23.86 10.24 11 41 30 3.87 rebounds 2 7 8.00 2.45 6 13 7 0.93
Example 2: Create Summary Table, Grouped by Specific Variable
The following code shows how to use the describeBy() function to create a summary table for the data frame, grouped by the ‘team’ variable:
#create summary table, grouped by 'team' variable describeBy(df, group=df$team, fast=TRUE) Descriptive statistics by group group: A vars n mean sd min max range se team 1 2 NaN NA Inf -Inf -Inf NA points 2 2 18.5 4.95 15 22 7 3.5 rebounds 3 2 7.5 0.71 7 8 1 0.5 steals 4 2 1.0 0.00 1 1 0 0.0 ------------------------------------------------------------ group: B vars n mean sd min max range se team 1 2 NaN NA Inf -Inf -Inf NA points 2 2 35.0 8.49 29 41 12 6.0 rebounds 3 2 6.0 0.00 6 6 0 0.0 steals 4 2 2.5 0.71 2 3 1 0.5 ------------------------------------------------------------ group: C vars n mean sd min max range se team 1 3 NaN NA Inf -Inf -Inf NA points 2 3 20.00 9.54 11 30 19 5.51 rebounds 3 3 9.67 3.06 7 13 6 1.76 steals 4 3 5.67 1.15 5 7 2 0.67
The output shows the summary statistics for each of the three teams in the data frame.
Additional Resources
How to Calculate Five Number Summary in R
How to Calculate the Mean by Group in R
How to Calculate the Sum by Group in R
How to Calculate Variance in R
How to Create a Covariance Matrix in R