Restore an ARDS dataset

The restore_ards function restores the ARDS data to wide format. The wide data can then be used for reporting.

Usage

restore_ards(data, init_vars = FALSE, anal_var = "anal_var")

Arguments

data

The input dataset to restore. The input dataset should correspond to the CDISC ARDS structure, such as that created by get_ards. However, not all variables are required. The only required variables are "anal_var", "statname", and "statval". All other variables will be processed if available, and ignored otherwise.

init_vars

Whether or not to keep the initialization variables on the restored data frames. Default is FALSE. The initialization variables include "studyid", "tableid", "adsns", "population", "time", and "where". To keep these variables on the restored data frames, set init_vars to TRUE.

anal_var

The name to use for the analysis variable column. This column is retained to positively identify the data frame. The default name is is "anal_var". If you need a different name for this column, pass the name as a quoted string. To eliminate the column entirely, pass a NULL value to this parameter.

Value

A list of data frames, transposed back into wide format. The list will have one or more items, distinguished by the analysis variable. The list item name will correspond to the name of the analysis variable.

Details

The init_ards, add_ards, and get_ards functions take data in wide format, and convert it to narrow format. The restore_ards function converts the narrow ARDS data back to wide format.

Wide format means there is a column for each statistic. Narrow format means all statistics are in a single column.

Because each analysis variable can have any number of statistics, when converting from narrow to wide, the resulting data frames can have different numbers of colums and different column names. Therefore, the restore_ards function returns a list of data frames, one for each analysis variable.

For each data frame, the statistics will each be in a separate column, named according to the original statistic variable name. The label of the statistics columns will be any value passed to the statistic description ("statdesc") for that analysis variable.

By default, the columns populated by init_ards will not be returned. These columns can be returned by setting the "init_vars" parameter to TRUE.

Once the ARDS data is restored and the statistics are back in separate columns, it will be easy to create a report, figure, or other output.

See also

Other ards: add_ards(), get_ards(), init_ards()

Examples

library(ards)
library(dplyr)

# Initialize the ARDS
# - These values will be common through the dataset
init_ards(studyid = "IRIS",
          tableid = "01", adsns = "iris",
          population = "all flowers",
          time = "1973")

# Perform analysis on Petal.Length
# - Using Species as a by-group
analdf1 <- iris |> 
  select(Petal.Length, Species) |> 
  group_by(Species) |> 
  summarize(n = n(),
            mean = mean(Petal.Length),
            std = sd(Petal.Length),
            min = min(Petal.Length),
            max = max(Petal.Length)) |> 
  add_ards(statvars = c("n", "mean", "std", "min", "max"),
           statdesc = c("Count", "Mean", "STD", "Minimum", "Maximum"),
           anal_var = "Petal.Length", trtvar = "Species")
           
# Perform analysis on Petal.Width
# - Using Species as a by-group
analdf2 <- iris |> 
  select(Petal.Width, Species) |> 
  group_by(Species) |> 
  summarize(n = n(),
            mean = mean(Petal.Width),
            std = sd(Petal.Width),
            min = min(Petal.Width),
            max = max(Petal.Width)) |> 
  add_ards(statvars = c("n", "mean", "std", "min", "max"),
           statdesc = c("Count", "Mean", "STD", "Minimum", "Maximum"),
           anal_var = "Petal.Width", trtvar = "Species")

# Get the ARDS
ards <- get_ards() 

# Convert back to wide format
res <- restore_ards(ards)

# View list names
print(names(res))
# [1] "Petal.Length" "Petal.Width" 

# Pull out Petal.Length
r1 <- res$Petal.Length

# View column names on Petal.Length
print(names(r1))
# [1] "Species"  "anal_var" "n"        "mean"     "std"      "min"      "max"    

# View stat data on Petal.Length
print(r1)
#      Species     anal_var  n  mean       std min max
# 1     setosa Petal.Length 50 1.462 0.1736640 1.0 1.9
# 2 versicolor Petal.Length 50 4.260 0.4699110 3.0 5.1
# 3  virginica Petal.Length 50 5.552 0.5518947 4.5 6.9

# Uncomment to view restored datasets
# View(res$Petal.Length)
# View(res$Petal.Width)