Home / UI Guide / Toolbar and Sidebar
Toolbar and Sidebar¶
Toolbar¶
The toolbar provides quick access to common operations:
File Operations¶
| Button | Shortcut | Description |
|---|---|---|
| New | Ctrl+N | Create a blank model |
| Open | Ctrl+O | Open a .cahtt or .pict model file |
| Save | Ctrl+S | Save the current model |
| Save As | Ctrl+Shift+S | Save with a new filename |
| Import PICT | Ctrl+I | Import a .pict format model |
| Export PICT | — | Export the model in .pict format |
Generation¶
| Button | Shortcut | Description |
|---|---|---|
| Generate | Ctrl+G or F5 | Generate the covering array |
| Stop | Escape | Stop an in-progress optimization |
Other¶
| Button | Description |
|---|---|
| Help | Open ProTest help |
| Advanced Mode | Toggle advanced tabs on/off (see below) |
Advanced Mode¶
The Advanced Mode checkbox in the toolbar controls the visibility of three additional tabs:
| Tab | What It Shows |
|---|---|
| Verification | Post-generation verification results — tuple coverage, constraint satisfaction, seed inclusion, and transitive constraint analysis |
| Trial Results | Detailed breakdown of each parallel trial — seed, test case count, timing, and which trial won |
| Model Text | Raw PICT model text with bidirectional sync to the visual editors |
When Advanced Mode is unchecked (the default), these tabs are hidden to keep the interface simple. The five standard tabs (Parameters, Constraints, Seeds, Sub-Models, Results) are always visible.
Left Sidebar¶
Parameters List¶
Displays all parameters in your model with their value counts. Clicking a parameter navigates to it in the Parameters tab.
Generation Controls¶
| Setting | Description |
|---|---|
| Order | Interaction strength (1-6). Default: 2 (pairwise) |
| Randomize | Enable randomized generation for varied results |
| Seed | Specific random seed for reproducible results (disabled when Randomize is off) |
| Case-sensitive | Enable case-sensitive parameter and value matching |
Algorithm¶
Select which generation engine to use:
| Engine | Description |
|---|---|
| PICT | Microsoft's greedy pairwise testing algorithm. Fast and reliable. |
| SIPO | Simulated annealing algorithm. Almost always produces fewer test cases than PICT — typically 10-30% fewer, though results vary by model. Recommended when minimizing test count matters. |
When SIPO is selected, additional tuning parameters become available:
| Setting | Default | Direction | Description |
|---|---|---|---|
| Base | 10,000 | Higher = fewer test cases, longer runtime | Number of optimization iterations per trial. Increasing the base gives SIPO more iterations to search for a smaller covering array, but each trial takes proportionally longer. |
| T-Init (Initial Temperature) | 5.0 | Higher = broader exploration | Starting temperature for simulated annealing. A higher T-Init lets the algorithm accept worse solutions early on, helping it escape local minima and explore a wider range of possibilities. Too high wastes iterations on poor solutions. |
| T-Final (Final Temperature) | 0.1 | Lower = finer refinement | Ending temperature. A lower T-Final makes the algorithm pickier in its final iterations, honing in on the best solution found. Too low and it may get stuck in a local minimum near the end. |
Parallel Trials¶
| Setting | Default | Direction | Description |
|---|---|---|---|
| Trials | 1,000 (SIPO) / 10,000 (PICT) | More = fewer test cases, longer runtime | Number of independent generation attempts run in parallel. Each trial uses a different random seed, and the smallest result is kept. More trials increase the chance of finding a smaller covering array, but total generation time increases. |
| Cores | ~75% of available CPU cores | More = faster generation | Number of parallel threads used for generation. The default leaves headroom so your system stays responsive. |
Validation Summary¶
Displays real-time validation status for your model. A green checkmark means the model is valid and ready to generate. Error and warning counts appear when issues are detected — errors block generation, warnings do not. Click individual messages to navigate directly to the source of the issue.
Validation runs automatically as you edit. Common issues flagged include:
- Empty or duplicate parameter names
- Duplicate values within a parameter
- Constraint syntax errors or references to unknown parameters/values
- Seeds referencing values that no longer exist
- Sub-models with fewer than 2 parameters or order exceeding member count