Cloud Collaboration with Crunch

Crunch.io provides a cloud-based data store and analytic engine. It has a web client for interactive data exploration and visualization. The crunch package for R allows analysts to interact with and manipulate Crunch datasets from within R. Importantly, this allows technical researchers to collaborate naturally with team members, managers, and clients who prefer a point-and-click interface: because all connect to the same dataset in the cloud, there is no need to email files back and forth continually to share results.

Subscribe to the mailing list to receive notification of releases and to ask general support questions.

Installing

crunch can be installed from CRAN with

install.packages("crunch")

The pre-release version of the package can be pulled from GitHub using the remotes package (part of devtools):

# install.packages("remotes")
remotes::install_github("Crunch-io/rcrunch")

Getting started

Connecting to Crunch and working with datasets is simple: load the crunch package and then login().

$ R
> library(crunch)
> login("jane.r_user@crunch.io")
Crunch password for jane.r_user@crunch.io:

You are now logged into Crunch as jane.r_user@crunch.io
[crunch] >
...

If you use OAuth to log into the web application, you’ll need to create a Crunch password to use with the R package. To set one, load the package, call resetPassword("your.email@example.com"), and check your email for instructions. You may need to log out of your current web app session before you click the link in that email to set your Crunch password.

Once you’ve logged in, check out listDatasets() to see the names of your existing datasets, which you can load like ds <- loadDataset("The name of my dataset"). New Crunch datasets can be made from a data.frame, a .csv or .sav file, or a URL to a file with newDataset(). See the help for these functions or vignette("crunch") for more information.

Additional configuration

You can set several parameters in your .Rprofile to simplify your workflow:

  • crunch.email and crunch.pw: you can save your Crunch credentials so that you don’t have to enter them each time. I.e., you can just > login(). Please be advised of the risks of storing your password like this. See ?login for more information. Also, note that you can opt to store just your crunch.email and enter your password each time you log in–a mix of convenience and security.
  • crunch.api: if not defined, it defaults to “https://app.crunch.io/api/”, the production server. To specify a different location, either set it in your .Rprofile or after loading the crunch package, do > options(crunch.api="https://otherapi.crunch.io/api/").

If you prefer, you can set the environment variables R_CRUNCH_EMAIL, R_CRUNCH_PW, and R_CRUNCH_API.

For developers

crunch requires R version 3.0 or greater. Note that if you’re on Ubuntu, particularly older versions, you’ll need to add an alternative PPA before trying to apt-get install R. For 12.04 Precise, for example,

# apt-key adv --keyserver keyserver.ubuntu.com --recv-keys E084DAB9
# echo "deb http://cran.r-project.org/bin/linux/ubuntu precise/" >> /etc/apt/sources.list
# apt-get update
# apt-get install libcurl4-openssl-dev
# sudo apt-get install r-base r-base-dev

See https://cran.r-project.org/bin/linux/ubuntu/README for more details.

Mocks

The tests for the rcrunch package relies on mocks that are compressed into a tarball so that the package stays within the CRAN 5MB size limit. To avoid git merge conflicts this file is not checked in so it must be created before you can run tests. It is created for you if you run make test, or you can create it explicitly by running make compress-fixtures or Rscript dev-misc/compress-fixtures.R.

Contributing

See the contribution guidelines.

Installing from a local checkout

# make deps
# make test

This installs dependencies and then runs the tests, which installs crunch from your local checkout in the process. If the dependencies fail to install, check the error message. You may need to install libcurl on your system before installing the R packages.

Running tests

You can run tests in an interactive session, or from the command line, $ make test is all you need. Requires the httptest package for R. You can also specify a specific test file or files to run by adding a “file=” argument, like $ make test file=auth. test_package will do a regular-expression pattern match within the file names. See its documentation in the testthat package.

Testing has two options: unit tests only, and tests that communicate with an API server. This is governed by an environment variable, INTEGRATION, which is false by default, meaning that API integration tests are not run. To modify this and test against the Crunch API, you can run $ make test INTEGRATION=TRUE.

To run integration tests, you will need to specify a test user, password, and API server to communicate with, either by setting test.user, test.pw, and test.api in your .Rprofile, as in:

options(
    test.user="magic.testuser@crunch.io",
    test.pw="t0pSecretP@ssw0rD",
    test.api="https://app.crunch.io/api/"
)

or by setting the environment variables R_TEST_USER, R_TEST_PW, and R_TEST_API.

If you are a Crunch developer serving a version of the API/backend with Vagrant or Docker, you will have best results if your R_TEST_API/test.api (1) is local.crunch.io, thanks to a mapping of localhost to that in your hosts file. In order to avoid self-signed certificate errors use the environment variable R_TEST_VERIFY_SSL=FALSE. You might point at “https://local.crunch.io:8443/api/”, for example. Some tests that cannot run successfully in the local environment will be skipped when run against a local.crunch.io URL.

Example of local usage:

$ R_TEST_VERIFY_SSL=TRUE R_TEST_USER=magic.testuser@crunch.io R_TEST_PW=t0pSecretP@ssw0rD R_TEST_API=https://local.crunch.io:28443/api/ make test INTEGRATION=TRUE file=variable-summary

Updating documentation

Run $ make doc. Requires the roxygen2 package.