Create a dataset

更新时间:
复制 MD 格式

A dataset is a collection of business data records, such as numbers, text, or dates. In Quick BI, datasets are the foundation for visual analysis. You can consolidate data from different data tables into one dataset to manage, analyze, and find valuable insights. Quick BI lets you create datasets either visually or with custom SQL. This topic describes how to create a dataset.

Prerequisites

You have connected to a data source. For more information, see Connect to a data source.

Procedure

  1. On the Quick BI homepage, navigate to the dataset creation page in one of the following ways:

    Entry 1: Create a dataset from the resource entry outside a workspace.

    image.png

    Entry 2: Quickly create a dataset from the resource list within a workspace.

    image.png

    Entry 3: Create a dataset on the Datasets page.image.png

    Entry 4: On the Data Sources page, click the image icon next to the target data table to create a dataset.image.pngEntry 5: On the Data Sources page, click Ad Hoc Query.image.png

  2. On the dataset creation page, select a data source (①) in the pane on the left. Then, drag a target data table (②) to the canvas on the right. Alternatively, create a dataset using custom SQL (③). For detailed instructions, see custom SQL.

    Note

    A dataset can contain up to 100 data tables or custom SQL queries.

    image

Next steps

  • Define relationships between multiple data tables to build a relationship model. You can also build a physical-layer model by using physical joins or merges. For detailed instructions, see Build a model.

  • After data modeling, you can create calculated fields and group dimensions. For detailed instructions, see data processing.

  • Configure settings for the dataset, such as the Quick Engine and permission control, to improve its performance. For detailed instructions, see advanced settings.

  • Manage created datasets. For detailed instructions, see Manage datasets.

  • Combine existing datasets based on your business requirements. For detailed instructions, see Combine datasets.

  • For additional information about combining datasets, see the Dataset FAQ.