dataMaid is an R package for documenting and creating reports on data cleanliness.
This github page contains the development version of dataMaid. For the latest stable version download the package from CRAN directly using
install.packages("dataMaid")
To install the development version of dataMaid run the following
commands from within R (requires that the devtools package is already installed)
devtools::install_github("ekstroem/dataMaid")
A super simple way to get started is to load the package and use the
makeDataReport() function on a data frame (if you try to generate several
reports for the same data, then it may be necessary to add the replace=TRUE
argument to overwrite the existing report).
library("dataMaid")
data(trees)
makeDataReport(trees)
This will create a report with summaries and error checks for each
variable in the trees data frame. The format of the report depends on your OS and whether
you have have a LaTeX installation on your computer, which
is needed for creating pdf reports.
The dataMaid package can also be used interactively by running checks for the individual variables or for all variables in the dataset
data(toyData)
check(toyData$events) # Individual check of events
check(toyData) # Check all variables at once
By default the standard battery of tests is run depending on the
variable type. If we just want a specific test for, say, a numeric
variable then we can specify that. All available checks can be viewed
by calling allCheckFunctions(). See the
documentation
for an overview of the checks available or how to create and include
your own tests.
check(toyData$events, checks = setChecks(numeric = "identifyMissing"))
We can also access the graphics or summary tables that are produced for a variable by calling the visualize or summarize functions. One can visualize a single variable or a full dataset:
#Visualize a variable
visualize(toyData$events)
#Visualize a dataset
visualize(toyData)
The same is true for summaries. Note also that the choice of checks/visualizations/summaries are customizable:
#Summarize a variable with default settings:
summarize(toyData$events)
#Summarize a variable with user-specified settings:
summarize(toyData$events, summaries = setSummaries(all = c("centralValue", "minMax"))
You can read the main paper accompanying the package at the Journal of Statistical Software. It provides a detailed introduction to the dataMaid package.
We also have two blog posts that provide an introduction to the package. The can be found here (the primary one) and here.
Moreover, we have
created a vignette that describes how to extend dataMaid to include
user-defined data screening checks, summaries and visualizations. This
vignette is called extending_dataMaid:
vignette("extending_dataMaid")
We are currently working on an online version of the tool, where users can upload their data and get a report. A prototype is already up and running - we just need to configure the R server correctly.
Until we have set it up online, you can try it out on your own machine:
library(shiny)
runUrl("https://github.com/ekstroem/dataMaid/raw/master/app/app.zip")