Create & manage datasets
Datasets are a core component of Redivis. They are a versioned collection of tables containing data, alongside rich documentation and metadata.
Datasets can be hosted by organizations or individual users, and every dataset has its own Dataset page. Datasets can be shared with other users on Redivis according to their access configuration.
Administrators can create datasets for their organization from the Datasets tab of their organization's Administrator panel. These datasets can be seen and managed by any administrator in the organization. When released they will be visible on the organization’s home page to anyone who as overview access tot he dataset.
Alternatively, anyone with a Redivis account can create a dataset on the Datasets tab of their Workspace. These datasets are by default only visible to their owner, and have simplified options to support sharing with your collaborators.
When you first create a dataset, it will be unpublished and only visible to other editors. This means you can edit the dataset and validate everything before releasing it. You can also validate and reconfigure its access rules before release.
At the core of every dataset is the data it contains, so we recommend starting here.
All data in a dataset is stored in tables. You can create a new table on the Tables tab of your dataset and start importing data. Redivis can upload data from your computer or another location you’ve linked, such as Box, Google Drive, AWS, and Google Cloud.
Once your data is finished importing you can validate that this table looks as you expect it to.
You can create more tables here if this dataset has multiple separate tables.
However, if your data is split across multiple files that all follow the same structure (such as a different file for each state, or each year of data but with generally the same variables) you will want to import all of these files to the same table, where they will be automatically appended together.
Metadata is essential to helping your researchers find and utilize your dataset. While some metadata will be generated automatically, such as variable summary statistics and counts, other metadata will require additional input.
On the Overview tab of the dataset there are multiple suggested sections to help break down information. You can fill out any of these that apply, such as the methodology, tags, contact information, etc.
Redivis will automatically generate citation and provenance information based on what we know about the dataset, but you can update this information with anything more specific.
If you have additional information that you want to include that doesn't fit one of these headers, you can create a custom section. Custom sections can also be set to be visible only to certain access levels if you have sensitive information.
You should also populate the metadata on each table. Tables can have a description, as well as an entity field that defines what each row in the table represents. You can also define the temporal and geographic range on the table, when relevant.
Each variable within a table has its own metadata. The variable name and type will be pre-determined from your data, but you should add a short label and longer description to each variable to help researchers understand what that variable measures.
Additionally, some variables will contain coded values, in which case you should provide value labels that represent the human-readable term for each code.
Once you are ready to make your dataset available to others, you'll need to release it. You can click the Review and publish or Review and release button in the top right of the page.
You should also confirm your access settings you set up when creating this dataset.
Once this dataset is released, it will become visible and available to anyone who would qualify for access.
You can use tools right on Redivis to create new versions of your dataset.
Administer your organization
Organizations allow for groups, centers, and institutions to more easily work with data by providing administrators with tools to effectively version and distribute data from a central location.