You are currently browsing the monthly archive for August 2015.

Five months ago we launched shinyapps.io. Since then, more than 25,000 accounts have been created and countless Shiny applications have been deployed. It’s incredibly exciting to see!

It’s also given us lots of data and feedback on how we can make shinyapps.io better. Today, we’re happy to tell you about some changes to our subscription Plans that we hope will make shinyapps.io an even better experience for Shiny developers and their application users.

New Starter Plan – More active hours and apps, less money
For many people the price difference between the Free and the Basic plan was too much. We heard you. Effective today there is a new Starter Plan for only $9 per month or $100 per year. The Starter Plan has the same features as the Free plan but allows 100 active hours per month and up to 25 applications. It’s perfect for the active Shiny developer on a budget!

More Active Hours for Basic, Standard, and Professional Plans
Once you’re up and running with Shiny we want to make sure even the most prolific developers and popular applications have the active hours they need. Today we’re doubling the number of active hours per month for the Basic (now 500), Standard (now 2,000), and Professional (now 10,000) plans. In practice, very few accounts exceeded the old limits for these plans but now you can be sure your needs are covered.

New Performance Boost features for the Basic Plan
In addition to supporting multiple R worker processes per application, which keeps your application responsive as more people use it, we’ve added more memory (up to 8GB) on Basic plans and above. While the data shows that most applications work fine without these enhancements, if you expect many users at the same time or your application is memory or CPU intensive, the Basic Plan has the performance boost you need. The Basic plan also allows unlimited applications and 500 active hours per month.

Traditionally, the mechanisms for obtaining R and related software have used standard HTTP connections. This isn’t ideal though, as without a secure (HTTPS) connection there is less assurance that you are downloading code from a legitimate source rather than from another server posing as one.

Recently there have been a number of changes that make it easier to use HTTPS for installing R, RStudio, and packages from CRAN:

  1. Downloads of R from the main CRAN website now use HTTPS;

  2. Downloads of RStudio from our website now use HTTPS; and

  3. It is now possible to install packages from CRAN over HTTPS.

There are a number of ways to ensure that installation of packages from CRAN are performed using HTTPS. The most recent version of R  (v3.2.2) makes this the default behavior. The most recent version of RStudio (v0.99.473) also attempts to configure secure downloads from CRAN by default (even for older versions of R). Finally, any version of R or RStudio can use secure HTTPS downloads by making some configuration changes as described in the Secure Package Downloads for R article in our Knowledge Base.

Configuring Secure Connections to CRAN

While the simplest way to ensure secure connections to CRAN is to run the updated versions mentioned above, it’s important to note that it is not necessary to upgrade R or RStudio to achieve this end. Rather, two configuration changes can be made:

  1. The R download.file.method option needs to specify a method that is capable of HTTPS; and
  2. The CRAN mirror you are using must be capable of HTTPS connections (not all of them are).

The specifics of the required changes for various products, platforms, and versions of R are described in-depth in the Secure Package Downloads for R article in our Knowledge Base.

Recommendations for RStudio Users

We’ve made several changes to RStudio IDE to ensure that HTTPS connections are used throughout the product:

  1. The default download.file.method option is set to an HTTPS compatible method (with a warning displayed if a secure method can’t be set);
  2. The configured CRAN mirror is tested for HTTPS compatibility and a warning is displayed if the mirror doesn’t support HTTPS;
  3. HTTPS is used for user selection of a non-default CRAN mirror;
  4. HTTPS is used for in-product documentation links;
  5. HTTPS is used when checking for updated versions of RStudio (applies to desktop version only); and
  6. HTTPS is used when downloading Rtools (applies to desktop version only).

If you are running RStudio on the desktop we strongly recommend that you update to the latest version (v0.99.473).

Recommendations for Server Administrators

If you are running RStudio Server it’s possible to make the most important security enhancements by changing your configuration rather than updating to a new version. The Secure Package Downloads for R article in our Knowledge Base provides documentation on how do this.

In this case in-product documentation links and user selection of a non-default CRAN mirror will continue to use HTTP rather than HTTPS however these are less pressing concerns than CRAN package installation. If you’d like these functions to also be performed over HTTPS then you should upgrade your server to the latest version of RStudio.

If you are running Shiny Server we recommend that you modify your configuration to support HTTPS package downloads as described in the Secure Package Downloads for R article.

The Joint Statistics Meetings starting August 8 is the biggest meetup for statisticians in the world. Navigating the sheer quantity of interesting talks is challenging – there can be up to 50 sessions going on at a time!

To prepare for Seattle, we asked RStudio’s Chief Data Scientist Hadley Wickham for his top session picks. Here are 9 talks, admittedly biased towards R, graphics, and education, that really interested him and might interest you, too.

Check out these talks

Undergraduate Curriculum: The Pathway to Sustainable Growth in Our Discipline
Sunday, 1400-1550, CC-607
Statistics with Computing in the Evolving Undergraduate Curriculum
Sunday 1600-1750
In these back to back sessions, learn how statisticians are rising to the challenge of teaching computing and big(ger) data.

Recent Advances in Interactive Graphics for Data Analysis
Monday,1030-1220, CC-608
This is an exciting session discussing innovations in interactive visualisation, and it’s telling that all of them connect with R. Hadley will be speaking about ggvis in this session.

Preparing Students to Work in Industry
Monday, 1400-1550, CC-4C4
If you’re a student about to graduate, we bet there will be some interesting discussion for you here.

Stat computing and graphics mixer
Monday, 1800-2000, S-Ravenna
This is advertised as a business meeting, but don’t be confused. It’s the premier social event for anyone interested in computing or visualisation!

Statistical Computing and Graphics Student Paper Competition
Tuesday, 0830-1020, CC-308
Hear this year’s winners of the student paper award talk about visualising phylogenetic data, multiple change point detection, capture-recapture data and teaching intro stat with R. All four talks come with accompanying R packages!

The Statistics Identity Crisis: Are We Really Data Scientists?
Tuesday, 0830-1020, CC-609
This session, organised by Jeffrey Leek, features an all-star cast of Alyssa Frazee, Chris Volinsky, Lance Waller, and Jenny Bryan.

Doing Good with Data Viz
Wednesday, 0830-1020, CC-2B
Hear Jake Porway, Patrick Ball, and Dino Citraro talk about using data to do good. Of all the sessions, you shouldn’t miss, this is the one you really shouldn’t miss. (But unfortunately it conflicts with another great session – you have difficult choices ahead.)

Statistics at Scale: Applications from Tech Companies
Wednesday, 0830-1020, CC-204
Hilary Parker has organised a fantastic session where you’ll learn how companies like Etsy, Microsoft, and Facebook do statistics at scale. Get there early because this session is going to be PACKED!

There are hundreds of sessions at the JSM, so no doubt we’ve missed other great ones. If you think we’ve missed a “don’t miss” session, please add it to the comments so others can find it.

Visit RStudio at booth #435

In between session times you’ll find the RStudio team hanging out at booth #435 in the expo center (at the far back on the right). Please stop by and say hi! We’ll have stickers to show your R pride and printed copies of many of our cheatsheets.

See you in Seattle!