Result Analyzer¶

The Result Analyzer performs statistical analysis across multiple test runs. It aggregates expectation values from all reports in logs/pass/ and logs/fail/, computes statistics (mean, standard deviation, Pp, Ppk), and renders histograms — helping you identify measurement drift, process capability issues, and intermittent failures.

Hotkey: F5

Overview¶

While the Result Viewer shows a single run's pass/fail results, the Result Analyzer looks at trends over time. It answers questions like:

Is a measurement drifting toward its limit?
What is the process capability (Pp/Ppk) for a given expectation?
Which tests have the lowest yield?
How spread are the measured values?

Workflow¶

Open the Result Analyzer (F5).
Configure filters (optional) — narrow down to specific serials, UUTs, sequences, or tests.
Click "Generate" — the analyzer scans all report files and computes statistics.
Review results — browse the hierarchical view with statistics and histograms.

Before generating, you can narrow the analysis scope:

Filter	Description
Serial Numbers	Only analyze reports with these serials. Empty = all.
Locations (UUTs)	Only analyze specific UUT positions. Empty = all.
Sequences	Only include these sequences. Empty = all.
Tests	Only include these tests. Empty = all.

Combine All Locations¶

When checked (default): all UUT positions are combined into a single statistical pool.

When unchecked: statistics are computed per-UUT position separately. Useful when different UUT slots have different measurement characteristics (e.g., different routing, different tolerances).

Statistics Computed¶

Per Test / Per Sequence¶

Statistic	Description
Total	Number of times this scope was executed
Passed	Number of times it passed
Pass %	Pass rate as a percentage
Enabled	How often it was enabled
Skipped	How often it was skipped (requirement unmet)
Average Duration	Mean execution time

Per Expectation (Numeric Values)¶

For expectations with numeric matchers (ToBeInRange, ToBeInPercentage, ToBeNear, ToBeGreater, ToBeLesser, etc.):

Statistic	Description
Total	Number of samples
Passed	Number that passed the expectation
Pass %	Yield for this expectation
Min Observed	Lowest measured value
Max Observed	Highest measured value
Mean	Average of all measured values
Median	Middle value (50th percentile)
Mode	Most frequently observed value
Range	Max − Min observed
Standard Deviation	Spread of values around the mean (lower is better)
Pp	Process Performance: `(USL − LSL) / (6σ)`
Ppk	Process Performance Index: `min((Mean − LSL) / 3σ, (USL − Mean) / 3σ)`

Where USL/LSL are the upper/lower specification limits from the expectation's matcher parameters (e.g., min and max for ToBeInRange).

Interpreting Pp/Ppk

Ppk ≥ 1.33 — process is capable with margin.
1.0 ≤ Ppk < 1.33 — process is capable but tight.
Ppk < 1.0 — process is not capable; failures are expected.
Ppk < 0 — the process mean is outside the specification limits.

Per Expectation (Exact Values)¶

For expectations with exact matchers (ToBeTrue, ToBeFalse, ToBeEqual, ToBeType):

Statistic	Description
Observed Values	Each unique value seen, with its occurrence count
Pass rate	How often the expected value was observed

Histogram¶

For numeric expectations, an interactive histogram is rendered using ImPlot:

X axis — measured values
Y axis — frequency (or density/cumulative, configurable)
Vertical lines — mark the specification limits (min/max)

Histogram Settings¶

Option	Description
Bin method	Sqrt, Sturges, Rice, Scott, or manual N bins
Horizontal	Flip the histogram orientation
Density	Normalize to probability density
Cumulative	Show cumulative distribution
No Outliers	Exclude outliers from bin calculation

Saving and Loading Reports¶

Save¶

After generating an analysis, click "Save Report" to export the results as a JSON file. This preserves the statistical summary for archival or sharing.

Load¶

Click "Load Reports" to open previously saved analysis files. Multiple files can be loaded simultaneously for comparison.

Results Hierarchy¶

The analysis results are displayed in a tree structure:

Location: UUT1
├── Overall: 95.2% pass (142/150), avg 4.2s
├── Power On (sequence)
│   ├── 100% pass, avg 0.8s
│   ├── Supply Voltage (test)
│   │   ├── 100% pass, avg 0.3s
│   │   └── VCC (expectation)
│   │       ├── Stats: mean=3.31, σ=0.02, Ppk=1.8
│   │       └── [Histogram]
│   └── Current Draw (test)
│       └── ...
└── Functional (sequence)
    └── ...

Tips¶

Run the analyzer after a production batch. After testing 50+ boards, the statistics become meaningful. Running it on 2-3 boards won't give useful Pp/Ppk values.
Watch for Ppk < 1.33. Even if everything is passing now, a low Ppk means you're at risk of future failures as the process drifts.
Use filters for focused analysis. If a specific test is failing intermittently, filter to just that test to see its value distribution clearly.
Compare locations. Uncheck "Combine all locations" to see if one UUT slot performs differently — this can indicate a fixture wiring issue.
Save reports before fixture changes. Save an analysis before and after a fixture modification to compare process capability.