CRAN downloads | dev build |
master build |
---|---|---|
dataCompareR is an R package that allows users to compare two datasets and view a report on the similarities and differences.
dataCompareR aims to make it easy to compare two tabular data objects in R. It’s specifically designed to show differences between two sets of data in a useful way that should make it easier to understand the differences, and if necessary, help you work out how to remedy them. In this regard, it aims to offer a more useful output than all.equal when your two datasets do not match, but isn’t intended to replace all.equal if you just want a binary test for equality.
rCompare()
does the comparison and creates a dataCompareR object containing all the differences between the two inputted datasets. The object can be used withprint
andsummary
.generateMismatchData()
generates a list of two data frames, each having the missing rows from the comparison.saveReport()
creates a summary of the comparison that is saved into a file.
It’s expected that dataCompareR will be used to compare data frames, but it can be used to compare any objects that can be coerced to data frames, such as data tables, tibbles or matrices. dataCompareR cannot compare data that is not tabular in format (nested JSON, irregular lists etc) but does handle tabular data that needs to be matched (or joined) on one or more keys (or ID columns).
Confirmed as working on R v3.6.3 and v4.0.0 for Windows, as well as v3.6.2, v4.0.0 and the devel
release for Linux.
Package was built with the following dependencies, but we anticipate it will work with later versions of these packages.
Package | Version | Source code URL |
---|---|---|
dplyr | 0.5.0 | https://github.com/hadley/dplyr |
knitr | 1.12.3 | https://github.com/yihui/knitr |
stringi | 1.0-1 | https://github.com/gagolews/stringi |
markdown | 0.7.7 | https://github.com/rstudio/markdown |
You can install from the CRAN via:
install.packages("dataCompareR")
You can also install the latest version directly from GitHub via
library(devtools)
install_git('https://github.com/capitalone/dataCompareR.git', branch = 'master',
subdir = 'dataCompareR', type = 'source', repos = NULL,
build_vignettes = TRUE)
Note: For OS X users: installing through install_git is broken. More information can be found here.
Please run vignette('dataCompareR')
after installation to see an example of the dataCompareR workflow.
The code is arranged as an R package, with the following contents:
- dataCompareR/R
- dataCompareR/man
- dataCompareR/tests/testthat
- dataCompareR/tests/performancetesting
- dataCompareR/inst/css
- dataCompareR/vignette
The contents will be covered below.
The main body of R code that provide the dataCompareR functionality.
The R package format mandates that this is a flat folder structure. Initial development had a nested structure, so to try to maintain this as far as possible, the naming convention for files is to preface them with 2 or 3 letter code that identifies the part of the code that file belongs to. The codes and hierarchy is as follows
- rc - rCompare - the entry point of the function
- pf - processFlow - handles the flow of an rCompare run
- vd - validateData - checks the data is suitable before starting an rCompare run
- pd - prepareData - prepares the input data for comparison
- cd - compareData - does the comparison
- rco - rCompare object - routines to handle the rCompare object that is generated by an rCompare run
- out - output - code to provide various views of the output
- pf - processFlow - handles the flow of an rCompare run
The filenames follow the format of the prefix, followed by underscore, followed by a camelcase description of what the code does. The .R files tend to have either 1 function inside them, or a small number of related functions.
Code is commented using ROxygen2 headers, which is used to automatically create the required R man pages by running
devtools::document()
Automated tests that are run via
devtools::test()
This consists of both unit tests and some end-to-end tests that MUST pass before any code is merged to dev or main. We've added Travis integration, so this is now mandated. If your development code change breaks an existing test, then it is your responsibility to fix it!
The current unit test coverage can be found in testing.md
- please feel free to add more tests, and regenerate this file using covR
.
This folder contains useful repeatable performance tests, but there are not run automatically, and the results they produce can only be interpreted manually.
https://cran.r-project.org/package=dataCompareR
- Version 0.1.0 released on 2017-07-17
- Version 0.1.1 released on 2017-11-14
- Version 0.1.2 released on 2019-09-07
- Version 0.1.3 released on 2020-05-01
We welcome and appreciate your contributions! Before we can accept any contributions, we ask that you please be sure to sign the Contributor License Agreement (CLA).
This project adheres to the Open Source Code of Conduct. By participating, you are expected to honor this code.