Second Partials Test Guide
The second partials test helps classify critical points of a two variable function. It is often used in calculus, optimization, economics, engineering, and statistics based modeling. A critical point is normally found where both first partial derivatives are zero. Once that point is known, the second partial derivatives describe local curvature near the point.
Why the Test Matters
The test gives a fast decision about local behavior. It can show whether a surface bends upward, bends downward, or changes direction. This is useful when a model has many possible input settings. A business model may need a maximum profit point. A loss function may need a minimum error point. A response surface may reveal saddle behavior.
Understanding the Determinant
The main value is the determinant D. It combines fxx, fyy, and fxy into one curvature measure. A positive determinant means the surface curves consistently in two main directions. Then fxx decides whether the point is a minimum or maximum. A negative determinant means the curvature changes direction. That creates a saddle point. A zero determinant gives no final answer.
Advanced Input Checks
This calculator includes optional first partial entries. These values help check whether the chosen point is truly critical. The tool also accepts fyx. That is useful when mixed partials are computed from separate work. If fxy and fyx differ, the page can flag that difference. You may average them when numerical estimates are being used.
Practical Interpretation
Always review the function and the point before trusting the result. The test is local only. It does not prove a global maximum or global minimum. Boundaries, constraints, and domain limits may change the final decision. Use the result as a classification step. Then compare it with graphing, constraint checks, or direct value testing when needed.
Using Results Carefully
Small determinant values can be sensitive to rounding. That is why tolerance is included. Increase tolerance when your derivatives come from measured data or numerical software. Use more decimals when your values are exact or highly precise.