Data view concepts - Amazon FinSpace

Data view concepts

Important

Amazon FinSpace Dataset Browser will be discontinued on November 29, 2024. Starting November 29, 2023, FinSpace will no longer accept the creation of new Dataset Browser environments. Customers using Amazon FinSpace with Managed Kdb Insights will not be affected. For more information, review the FAQ or contact AWS Support to assist with your transition.

View Types

Two types of views can be setup for a dataset:

  • Auto-Update view – A data view with all additions (Append) and corrections (Replace, Replace Changesets) for a dataset. Future additions and corrections to this dataset are automatically applied to this view.

  • Static view – A data view with all additions (Append) and corrections (Replace, Modify) up to a specified date and time for creation of view i.e. the view will be constructed from only those changesets that were created before the specified time and date. No future additions or corrections will be applied to this view.

A screenshot that shows the types of data views.

A data view is constructed from changesets. Two factors are taken into account for the changesets to be considered in a view:

  1. Specified date and time to create the view – All the changesets created prior to the specified date and time are considered for the view. In case of an auto-update view, the specified date and time is current day and timestamp.

  2. The changeset types are interpreted for a creating a data view in the following ways:

    • Changeset with Append type – Changeset is interpreted as an addition to the end of all the prior created changesets. The changeset will be considered for view creation.

    • Changeset with Replace type – Changeset is interpreted as a replacement to all prior created changesets. No changesets created before a changeset of this type are considered for the view creation.

    • Changeset created as a correction – Changeset is interpreted as a replacement to a specific prior created changeset. The prior created changeset will be not considered for the view creation.

View Last Updated

The timestamp represents the point in date and time for which the view is created. For static view, it will be the timestamp that you specified at the creation of the view. For auto-update view, it will be the last time it was updated when a new changeset was added.

Data View ID

The unique identifier for a data view.

Dataset ID

The unique identifier for a dataset.

Data Access

A view can be prepared to be accessed and used in:

  • FinSpace notebook using integrated Spark clusters.

  • Externally via the FinSpace API. The format of this view can be customized by specifying file format, delimiter, compression type.

Partitioning

Partitioning can be configured to optimize queries.

Sorting

The data in the data view can be sorted by one or more columns. Sorting data helps with query performance.