Query Environment Priority

Before You Start

Metric queries may involve data filters in multiple places.

To describe the query runtime environment more simply, this page treats grouped dimensions as filter dimensions.

The Sales Amount metric result in the example on the left can be understood as having three filters:

Metric Date = 2024-09-30

Province = Zhejiang

City = Hangzhou

With this understanding, the following descriptions of query environment calculation order are based on filters.

Query Environment

When a user queries a metric result in the interface, the metric definition may be nested across multiple layers. For example, a derived metric may reference a basic metric. The system therefore generates multiple nested query environments.

The following example uses a slightly complex metric definition:

Metric name: Average Monthly Inventory of Hot-Selling Products in Recent 12 Months

Type: derived metric

Base/composite metric: Inventory

Time limit: recent 12 months, monthly average

Business limit: Product ID satisfies Sales Volume in Recent 12 Months > 1000

Metric name: Inventory

Type: basic metric

Aggregation content: SUM([Inventory])

Semi-additive metric: enabled

Non-additive dimension: Date

Window selection: Maximum

Metric name: Sales Volume in Recent 12 Months

Type: derived metric

Base/composite metric: Sales Quantity

Time limit: recent 12 months

Metric name: Sales Quantity

Type: basic metric

Aggregation content: COUNT([Order ID])

The query conditions in the metric query are: Metric Date (Month) = 2024-09 and Warehouse Province = Zhejiang.

The following diagram shows the relationship between query environments:

Rules

Rules

1. The query environment is passed from outside to inside, and data calculation is performed from inside to outside.

2. Filter conditions in query environments at the same level do not affect each other.

Based on these rules, the calculation logic can be broken down into the following query environments:

Step	Environment	Processing Logic
1	1	Generate filters: - Metric Date (Month) = 2024-09 - Warehouse Province = Zhejiang Pass these filters to environment 2.
2	2.1	Receive the filters from environment 1 and merge them with its own filters: - Metric Date (Month) = 2024-09 - Warehouse Province = Zhejiang - Product ID satisfies Sales Volume in Recent 12 Months > 1000 Because a time limit exists, adjust the metric date passed from environment 1: - Metric Date (Month) = 2023-09 ~ 2024-09 - Warehouse Province = Zhejiang - Product ID satisfies Sales Volume in Recent 12 Months > 1000
	2.2	Calculating the filter Product ID satisfies Sales Volume in Recent 12 Months > 1000 uses the metric result of Sales Volume in Recent 12 Months. Receive filters from environment 1: - Metric Date (Month) = 2024-09 - Warehouse Province = Zhejiang Because a time limit exists, adjust the metric date passed from environment 1: - Metric Date (Month) = 2023-09 ~ 2024-09 - Warehouse Province = Zhejiang
3	3.1	Receive the filters from environment 2.1: - Metric Date (Month) = 2023-09 ~ 2024-09 - Warehouse Province = Zhejiang - Product ID satisfies Sales Volume in Recent 12 Months > 1000
3	3.2	Receive the filters from environment 2.2: - Metric Date (Month) = 2023-09 ~ 2024-09 - Warehouse Province = Zhejiang

The calculation runs through the following steps:

Using the query cell for **Metric Date (Month) = 2024-09** and **Warehouse Province = Zhejiang** as the example:

Calculate within environment [3.2]:

1. Use Product ID as the grouping dimension, calculate the result for Metric Date (Month) = 2023-09 ~ 2024-09 and Warehouse Province = Zhejiang, and obtain multiple rows of data.

Product ID	Sales Volume in Recent 12 Months

Calculate within environment [3.1]:

1. Use Metric Date (Month) as the grouping dimension to calculate the semi-additive metric within Metric Date (Month) = 2023-09 ~ 2024-09 and Warehouse Province = Zhejiang.

2. Because this is a semi-additive metric and the filter conditions contain the non-additive dimension Metric Date (Month), the metric first keeps the maximum date data within the Metric Date (Month) range.

3. On the filtered result from the previous step, filter Warehouse Province = Zhejiang and keep the Product ID values whose values in [3.2] are greater than 1000. This returns multiple rows of data from 2023-09 to 2024-09.

Metric Date (Month)	Inventory of Hot-Selling Products in Recent 12 Months

1. Calculate the average of the data result from step 3 to obtain the final result: Average Monthly Inventory of Hot-Selling Products in Recent 12 Months.