Overview

The "Datasets" section is a fundamental part of custom metrics, where users can access and manage various data essential for creating metrics, charts, and reports. This section serves as the foundation for data-driven decision-making and analysis within your system.

Within the "Datasets" section, users will find several main sections, each catering to a specific type of dataset:

1. Common Datasets: This section contains data sources that have been pre-configured in Perf, simplifying data access for users. Common datasets typically include Task Tracking Systems (TTS) and Version Control Systems. Users can seamlessly integrate these data sources into their metric-building processes.

2. Uploaded Datasets: In this section, users can access datasets that were manually uploaded in CSV format. These datasets provide flexibility for users to bring in custom data for analysis and reporting.

3. Custom Datasets: Custom datasets are data sources that are connected via API using custom data loaders. Users can tap into external systems and databases to bring in specific data relevant to their metrics. Custom datasets offer extensive flexibility and customization options.

4. Catalog Metric Data: This section houses data retrieved from the catalog metrics. Users can access pre-defined metrics and incorporate them into their analyses, streamlining the process of metric creation.

5. Manual Metrics Data: In this section, data from Metrics & KPIs (also known as Manual Metrics) is presented. Users can utilize this manual metrics data for calculation and building their own custom metrics, enhancing the overall project analysis for a more holistic and accurate understanding.

Structure and Features

Here are some important structural features of the "Datasets" section. Each of the main sections listed above is further divided into sub-sections, providing users with organized access to their datasets.

• Tables: Within each sub-section, users will find tables that represent specific datasets. These tables display key information and options for each dataset.

• Columns: Tables within the sub-sections include various columns, each designed to display specific data attributes or properties. These columns can include data types such as Array, Boolean, Date, JSON, Numeric, and Text, providing users with diverse data-handling capabilities.

Data Types

Each filed has an icon to display type of a data in them:

	Text
	Date
	Numeric
	Array
	Boolean
	JSON

Joins Between Tables

By default, Custom Metric v2 is created based on one table at a time. By using joins, you can retrieve data from two or more tables based on logical relationships between the tables and create a chart based on them.

SELECT * FROM table_A <JOIN_TYPE> table_B ON join_condition