Skip to main content

Custom Analysis

In "Custom Analysis" you can run Custom Data Analysis and Custom Data Management. Define query metrics and filter conditions to inspect data trends and details, or apply an existing template to define query metrics quickly. Use filters to control statistical toggles, and edit or create key-value records.

1. Custom Data Analysis​

1) Query Metrics​

Define query metrics here. For example:

  • Metric 1: Avg FPS per match → aggregate mean
  • Metric 2: Scene duration per match → aggregate mean

You can also click "Switch to Template Analysis" to apply an existing metric template directly. Alt text

2) Filter Conditions​

Edit the filters: time period, version, scene, CPU, GPU, device, quality settings, and so on. You can also use "Add metrics" to add other metrics (such as Avg FPS or FPS variance) as filters. Alt text

3) Performance Trend​

After clicking "Search", the Performance Trend module shows the trend of every metric. Switch metrics by clicking different tabs, and switch between granularity (hourly / daily) and trend types (time trend / day-over-day / week-over-week). Alt text

4) Data Details​

In Data Details you can switch the statistics dimension (day / hour / period) and grouping dimensions (version, scene, model, SoC, GPU, etc.; multi-select supported).

In the data table click "Drill Down" to drill the data, or click "View Match" to jump to the match detail list. Alt text

2. Custom Data Management​

Manage and review the custom data already reported. Set "Open statistics" to open/close statistics for each metric and control whether it can be selected in Custom Data Analysis. Alt text