RStudio Connect 1.8.0

January 21, 2020

RStudio Connect helps data science teams quickly make an impact by enabling them to publicize reports, models, dashboards, and applications created in R and Python with business stakeholders. The 1.8.0 release makes it even easier for teams to start sharing.

For Data Scientists, highlights include:

  • Python Support: Improvements that make it easier to share Jupyter Notebooks and mixed R and Python content.
  • pins: An easy way to share data, models, and other objects.
  • Custom Emails: Use code to create beautiful emails with plots and results inline.

For DevOps / IT Administrators, 1.8.0 makes it easier to support data science teams in production:

For Data Science Leaders, 1.8.0 simplifies onboarding new team members and eases collaboration:

To learn more about all the ways RStudio Connect makes it easy to connect your Data Science team with your decision makers, visit our website. An easy way to get started is with the RStudio Team Quickstart to experience all of RStudio’s products on an easy-to-use virtual machine, or begin a free 45-day evaluation of RStudio Connect.

Python Support

Publish Options in Jupyter

RStudio Connect has supported both R and Python for over a year, and during this time we’ve made significant improvements. Data scientists can now publish Jupyter Notebooks with a single click or by using version control. Data scientists who use both R and Python also have more flexibility, helping them deploy mixed content using the reticulate R package.

Publish a Jupyter Notebook to RStudio Connect

Pins

The pins package makes it easy to share data, models, and other objects on RStudio Connect. Pins are especially useful if you have data that regularly updates; simply schedule an R Markdown document to process your data and pin your results or model. Once pinned, your reports, applications, and APIs can automatically pull the updates. Learn more or see an example.

Pins Support in RStudio Connect

Custom Emails

Sending plots, tables, and results inline in an email is a powerful way for data scientists to make an impact. RStudio Connect customers use custom emails to send daily reminders, conditional alerts, and to track key metrics. The latest release of the blastula package makes it even easier for data scientists to specify these emails programmatically:

if (demand_forecast > 1000) {
  render_connect_email(input = "alert-supply-team-email.Rmd") %>%
  attach_connect_email(
    subject = sprintf("ALERT: Forecasted increase of %g units", increase),
    attach_output = TRUE,
    attachments = c("demand_forecast_data.csv")
  )
} else {
  suppress_scheduled_email() 
}
Custom Emails from RStudio Connect

Learn more here!

Jump Start Examples

A common challenge facing data science teams is onboarding new users. Data scientists have to learn new tools, methods, and often a new domain. We’ve created a set of examples to help data scientists learn common best practices. A new tutorial in the product helps users publish different data products, which is particularly helpful for data scientists exploring new content types, such as reports, models, or datasets.

Jump Start Examples

Git-centered Deployments

RStudio Connect’s push-button publishing is a convenient and simple way for data scientists to share their work. However, some teams prefer Git-centric workflows, especially when deploying content in production. RStudio Connect supports these workflows, making it effortless for data science teams to adopt version control best practices without maintaining additional infrastructure or remembering complex workflows. Data scientists simply commit to Git, and RStudio Connect will update the content, saving you any extra steps.

Create New Content from Git Repository in RStudio Connect

Scheduled Content Calendar

For system and application administrators, RStudio Connect makes it simple to audit data science work. For data science teams, one powerful application of RStudio Connect is the ability to schedule tasks. These tasks can be everything from simple ETL jobs to daily reports. In 1.8.0, we’ve made it easier for administrators to track these tasks across all publishers in a single place. This new view makes it possible to identify conflicts or times when the server is being overbooked.

View Scheduled Content

Security Updates & Deprecations

  • RStudio Connect no longer supports Ubuntu 14.04 LTS (Trusty Tahr).
  • The Postgres.URL database connection URL will always use a nonempty Postgres.Passwordvalue as its password. Previous releases would use the password only when a {$} placeholder was present. Support for the Postgres.URL placeholder {$} is deprecated and will be removed in a future release.
  • Duplicate user names are now accepted in some conditions for LDAP, SAML, and proxied authentication providers. See the [release notes] for details.
  • Support for TLS 1.3. Access to this TLS version is available without additional configuration.
  • The setting HTTPS.ExcludedCiphers has been removed and is no longer supported. The HTTPS.MinimumTLS setting should be used to specify a minimum accepted TLS version. We recommend running a secure proxy when your organization has more complex HTTPS requirements.
  • The deprecated setting Applications.DisabledProtocols has been removed. Use Applications.DisabledProtocol instead. Multiple values should be placed one per line with this new setting.
  • The deprecated settings OAuth2.AllowedDomains and OAuth2.AllowedEmails have been removed. Use OAuth2.AllowedDomain and OAuth2.AllowedEmail instead. Multiple values should be placed one per line with these new settings.
  • TensorFlow Model API deployment supports models created with TensorFlow up to version 1.15.0. A systems administrator will need to update the version of libtensorflow.so installed on your RStudio Connect system.

For more information about all the updates available in RStudio Connect 1.8.0, we recommend consulting the release posts for the 1.7 series.

This release also includes numerous bug fixes, the full release notes document all of the changes. Some of our favorites:

  • The “Info” panel for Jupyter notebook content includes a summary of recent usage.
  • Python-based content can now use an application RunAs setting.
  • Scheduled reports no longer run multiple times if the application and database clocks disagree, or in the case of concurrent rendering attempts in a multi-node installation.

Upgrade Planning

Aside from the deprecations and breaking changes listed above, there are no other special considerations, and upgrading should require less than five minutes. If you are upgrading from a version earlier than 1.7.8, be sure to consult the release notes for the intermediate releases, as well.

Get Started with RStudio Connect

If you haven’t yet had a chance to download and try RStudio Connect, we encourage you to do so. RStudio Connect is the best way to share all the work that you do in R and Python with collaborators, colleagues, or customers.

You can find more details or download a 45-day evaluation of the product at https://www.rstudio.com/products/connect/. Additional resources can be found below:

Previous Article
rstudio launcher is not working. Throwing error: [rserver] ERROR Cannot connect to the Job Launcher service
rstudio launcher is not working. Throwing error: [rserver] ERROR Cannot connect to the Job Launcher service

@diegop wrote: Hi I'm trying to setup launcher service in my rstudio server inst...

Next Article
RStudio Connect 1.8.0
RStudio Connect 1.8.0

RStudio Connect helps data science teams quickly make an impact by enabling them to publicize reports, mode...