Frequently asked questions

Compute Capsules work by creating an immutable record of the code, environment, and data used in computational work, as well as the results produced. They are releasable, versionable, and reproducible. Whatever runs today will always run in the future.

Virtually anything. Simple exploratory analysis, training machine learning models, creating API connections, building visualization apps, and more. However, you can do all these things at the same time as attaching data of any size, provisioning compute, and co-developing with others, all while creating an automatically traceable and reproducible record of work.

VSCode, JupyterLab, RStudio, CLI, MATLAB, Shiny, Streamlit, Ubuntu Desktop (Beta) and IGV (Beta).

Docker is just one piece of what makes up a Capsule. Docker is natively integrated into Code Ocean so users don’t need to be familiar with Docker to take advantage of its containerization technology. One example is the automated Dockerfile generation. 

Both. All Capsules (and Pipelines) on Code Ocean integrate Git to track changes as development progresses, regardless of whether or not you’ve connected Code Ocean to your Git provider. We have native integrations for Github, Gitlab, Bitbucket and Azure DevOps.

No. You can attach data of any size to any Capsule. Data and Capsules are managed separately, meaning you can attach the same data asset to multiple Capsules at the same time, reducing data duplication, and potentially decreasing the total cost of ownership over time.

Nextflow. Pipelines have native integration of nextflow, with nextflow scripts automatically created and written as uesrs work in the visual pipeline builder.

Both. Building Pipelines with Code Ocean’s Pipeline Builder UI uses DSL1. Custom pipelines support both DSL1 and DSL2.

No. Pipelines can be unlocked for users to write their own custom script in Nextflow. Users can fully leverage all Nextflow capabilities within a custom pipeline.

Yes. Each Capsule can have different computing resources applied (when part of a pipeline in the visual UI), depending on requirements.

Yes. Users can import an existing pipeline from a git repository, import from nf-core, or upload files to make their custom pipeline. This allows them to quickly create pipelines when first starting with Code Ocean.

Internal Data that you’ve brought into Code Ocean for enhanced reproducibility are stored in an organized S3 bucket inside your AWS account. External data remains external and is accessed by linking to S3 bucket locations. 

Virtually anywhere. You can bring in data by local upload, and native integrations include AWS S3, AWS Athena, and Databricks. Code Ocean Apps offers several custom Data Connector Capsules, or you can make your own by connecting through API.

No. There are no limits to the size of Data that Code Ocean supports, and any data asset can be connected to any Compute Capsule or Pipeline.

Yes. Code Ocean supports user Secrets for any credentials that are required to access private data and assumable IAM roles.

Yes. Data can be easily shared with users and groups. Secrets and Roles ensure data access is a seamless experience for users.

Yes. Admins can create custom keys and specify value and input types for users. Custom metadata allows for improved search, management, and organization of data.

Yes, if all source data has been imported and made immutable. If you’ve used an external data source that is subject to change, reproducibility cannot be guaranteed.

Yes. You can interact with every node on the Lineage Graph. Click on any asset that you have access to and explore the full details of which version was run, when it was run, and who by.

Yes, using our API, you could connect to your own instance of an open-source tool, e.g. OpenLineage.

You can add all four of the main computational assets in Code Ocean: Compute Capsules, Pipelines, no-code Apps, and Data.

The owner(s) of an asset can add or remove it from a Collection.

No. Every asset has its own access permissions. Users will only see what they have permission to see in a Collection.

Admins determine the Collections folder structure for the organization.

Capsules, Pipelines, and Apps that are not yet released/still in development cannot be added to a Collection. This makes sure that anything added to a collection has passed important checks.

There are three options. Streamlit and RShiny are both available as cloud workstations. Any Capsule or Pipeline can also be converted into an app with the app panel.

Yes. Any Capsule or Pipeline can be converted into a parameterized no-code App that can then be released for use by non-coders.

Yes. Users can easily swap out App data for their own in an App Panel. This is also possible with Streamlit and RShiny.

We often add new apps, usually at the request of existing customers. To find out if we can help you with this, please get in touch.

We support built-in authentication, SAML SSO, and OpenID Connect (OIDC).

An analytics dashboard allows administrators to monitor usage, costs, storage, and activity across their entire deployment. Admins are also able to get direct access to users' assets (there are hyperlinks that make is very easy for admins to identify and navigate to them).

Yes. Dedicated machines can be inactivated to prevent users from selecting these machines for use in their workflows.

Yes. Custom environments can be deployed or even created from a Compute Capsule. Admin-managed Starter Environments are a great way to standardize workflows across the organization.