Back to Details

visualize-data

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Goals

目标

  • Create efficient but highly readable tables/charts by applying good visualization settings and formattings.
  • Users might want to save the visualization for long-term use, so make it nice.
  • 通过应用合理的可视化设置与格式规范，创建高效且可读性极强的表格/图表。
  • 用户可能需要保存可视化内容长期使用，因此需保证输出质量美观。

Pre-requisites

前置条件

  • Set up Holistics MCP
  • 已搭建 Holistics MCP 环境

Recommended workflow

推荐工作流

  1. generate_aql
    to query the raw data (WITHOUT visualization requirements)
  2. generate_viz
    to create a viz on top of that AQL query (apply visualization requirements and default formattings)
  3. execute_viz
    to run the viz and show the result to the user
If 
generate_viz
fails, fall back to showing a plain table with 
execute_aql
.
  1. generate_aql
    用于查询原始数据（无需携带可视化需求）
  2. generate_viz
    用于基于上述AQL查询结果创建可视化内容（应用可视化需求与默认格式规范）
  3. execute_viz
    用于运行可视化任务并向用户展示结果
如果 
generate_viz
执行失败，回退到使用 
execute_aql
展示纯文本表格。

Rules

规则

  • To update date/time filters, use 
    generate_aql
    first. It can write accurate date/time condition syntax.
  • Cohort retention should ALWAYS be visualized as Pivot Table.
  • generate_aql
    knows how to calculate cohort retention. You don't need to explain the steps when using it.
  • Always use 
    generate_viz
    to apply proper display formats for new percentage and currency dimensions/metrics.
    • Percentage: Make sure to display with proper "%" format
    • Currency: Make sure to include the proper currency symbol
  • Review the result and check that it meets the visualization goals. Follow up to improve your visualizations (using 
    generate_viz
    and 
    execute_viz
    ) if needed.
  • 如需更新日期/时间筛选条件，请先调用 
    generate_aql
    ，它可以生成准确的日期/时间条件语法。
  • 群组留存（Cohort Retention）必须始终以透视表形式可视化。
  • generate_aql
    内置群组留存计算逻辑，调用时无需额外解释计算步骤。
  • 必须 使用 
    generate_viz
    为新增的百分比、货币类维度/指标应用正确的展示格式：
    • 百分比：确保以正确的带「%」格式展示
    • 货币：确保包含正确的货币符号
  • 复盘结果，检查是否满足可视化目标。如有需要，可通过 
    generate_viz
    和 
    execute_viz
    迭代优化可视化效果。

Tool: 
generate_viz

工具：
generate_viz

  • To visualize plain tables (without any formatting, colors, etc.), you don't need to call this. You can use 
    execute_aql
    directly.
  • To visualize non-plain tables (e.g. tables with conditional formatting, percentage formatting, colors, etc.) or any other visualization types that require AML visualization features → call this tool.
  • Unless the viz type is specifically requested by the user, let the tool decide the viz type itself.
  • 如需可视化纯表格（无任何格式、颜色等配置要求），无需调用此工具，可直接使用 
    execute_aql
    。
  • 如需可视化非纯表格（例如带条件格式、百分比格式、颜色配置等的表格）或其他需要使用AML可视化能力的可视化类型 → 调用此工具。
  • 除非用户明确指定可视化类型，否则由工具自行决定可视化类型。