Automated Regression Test Matrix

Test Suite Controls

Grid resolution (N×N cases) 6×6

Injected bug severity (drag coeff.) 0.10

Pass tolerance (RMSE, m) 0.50

Sweep speed (tests/s) 6

Live Results

Progress

0 / 36

Pass rate

—

Avg RMSE

— m

Current case

—

Pass Fail Pending

How it works

Each grid cell is one automated test case: a projectile is launched with speed v (columns) and angle θ (rows). The expected result is the closed-form, drag-free trajectory:

y_exp(x) = x·tanθ − g·x² / (2v²cos²θ)
R_exp    = v²·sin(2θ) / g

The simulated result — the "model under test" — is produced by numerically integrating Newton's second law with explicit Euler steps, including a quadratic drag term whose strength is the injected bug severity k:

ax = −k·|v|·vx,  ay = −g − k·|v|·vy
vx += ax·dt,  vy += ay·dt,  x += vx·dt,  y += vy·dt

For each test case, five checkpoints along the expected range are compared against the simulated trajectory (linearly interpolated to the same x), and the case's score is the Root Mean Square Error:

RMSE = √[ Σ(y_sim,i − y_exp,i)² / n ]

A case turns green when RMSE stays under the tolerance threshold and red when it exceeds it — exactly the pass/fail logic a CI pipeline applies when it regression-tests a physics model against a trusted baseline after every code change. Raising the bug slider simulates a real regression: at k=0 the numerical integrator tracks the analytic solution closely and nearly every case passes; as k grows, longer, flatter shots (large v, low θ, more drag exposure) fail first, since they spend the most time accumulating drag.

RUNNER: idle

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.