atableExtra

atable is a package simplifying the task of creating tables of summary statistics similar to those found as the first table of a clinical trial publication. Although it is very flexible, utilising this flexibility involves writing new functions to calculate other measures and statistics. atableExtra provides some additional functions to simplify this task. atableExtra also changes some atable default settings.

Installation

# install.packages("devtools")
devtools::install_github("aghaynes/atableExtra")

Factors

atableExtra changes the default formatting of factors from percentage (N) to N (percentage).

library(atable)
atable_options(format_to = "Console")
data(mtcars)
mtcars$cyl <- as.factor(mtcars$cyl)
atable(mtcars, target_cols = "cyl")
#>   Group         value   
#> 1 Observations          
#> 2               32      
#> 3 cyl                   
#> 4      4        34% (11)
#> 5      6        22% (7) 
#> 6      8        44% (14)
#> 7      missing  0% (0)

library(atableExtra)
#> Loading required package: glue
#> Registered S3 method overwritten by 'atableExtra':
#>   method                              from  
#>   format_statistics.statistics_factor atable
atable(mtcars, target_cols = "cyl")
#>   Group         value   
#> 1 Observations          
#> 2               32      
#> 3 cyl                   
#> 4      4        11 (34%)
#> 5      6        7 (22%) 
#> 6      8        14 (44%)
#> 7      missing  0 (0%)

Numeric

atableExtra also provides a flexible approach to formatting numeric variables via the numeric_stats function and it’s appropriate format_statistics method. It’s default settings mimic those from atable:

# atable
atable(mtcars, target_cols = "mpg")
#>   Group                value 
#> 1 Observations               
#> 2                      32    
#> 3 mpg                        
#> 4      Mean (SD)       20 (6)
#> 5      valid (missing) 32 (0)
# atableExtra
atable_options(statistics.numeric = numeric_stats) # replace the default function
atable(mtcars, target_cols = "mpg")
#>   Group                value 
#> 1 Observations               
#> 2                      32    
#> 3 mpg                        
#> 4      Mean (SD)       20 (6)
#> 5      Valid (missing) 32 (0)

It’s easy to change the format though via the numstats argument that takes either numbers 1 to 3 (defaults to 1)…

atable(mtcars, target_cols = "mpg", numstats = 1:3)
#>   Group                   value      
#> 1 Observations                       
#> 2                         32         
#> 3 mpg                                
#> 4      Mean (SD)          20 (6)     
#> 5      Min - Max          10 - 34    
#> 6      Median [Quartiles] 19 [15; 23]
#> 7      Valid (missing)    32 (0)

… or named characters…

atable(mtcars, target_cols = "mpg", 
       numstats = c("Mean" = "{mean}", "Median" = "{q0.5}"))
#>   Group                value 
#> 1 Observations               
#> 2                      32    
#> 3 mpg                        
#> 4      Mean            20    
#> 5      Median          19    
#> 6      Valid (missing) 32 (0)

Any quantile can be used (0th (= minimum), 25th, median, 75th and 100th (=maximum) are available by default) via the quantiles argument and referred to as e.g. q0.1 for the 10th percentile.

atable(mtcars, target_cols = "mpg", 
       quantiles = .1, 
       numstats = c("10th percentile" = "{q0.1}"))
#>   Group                value 
#> 1 Observations               
#> 2                      32    
#> 3 mpg                        
#> 4      10th percentile 14    
#> 5      Valid (missing) 32 (0)

The number of valid/missing observations can be turned off via the missingformat argument…

atable(mtcars, target_cols = "mpg", missingformat = FALSE)
#>   Group          value 
#> 1 Observations         
#> 2                32    
#> 3 mpg                  
#> 4      Mean (SD) 20 (6)

… or modified as per the numstats argument…

atable(mtcars, target_cols = "mpg", missingformat = c("Missing" = "{Nmissing}"))
#>   Group          value 
#> 1 Observations         
#> 2                32    
#> 3 mpg                  
#> 4      Mean (SD) 20 (6)
#> 5      Missing   0