Purdue MA 26100 · official Exam 1 study guide
The department's own study guide is the source of truth for what Exam 1 covers, and it gives you a bare answer under each problem with no work in between. This is that guide, section for section and problem for problem — with the work filled in, the traps named, and the answer kept shut until you commit to an attempt. All 68 of its live problems, plus one it leaves commented out.
Source of truth: the official Exam 1 study guide, read 2026-09-14. Every section heading, formula box and problem statement below is the instructor's, unedited. The worked steps, trap notes and corrections are mine. Two of the published answers are wrong — the page shows the official answer and the correction side by side. · Reset my progress · Exam 2 → · Final →
Everything below is the source of truth for the material. But the Fall lesson split is not identical to the Spring one, and the difference is examinable:
The date printed on the official guide is the Spring exam. Yours is Mon Oct 5 · 8:00 pm.
A vector has both magnitude (length) and direction. Vector notation:
where , , are the standard basis vectors.
Magnitude (Length) of vector :
A unit vector is a vector of length 1. The unit vector in the direction of is:
Dot Product of and :
where is the angle between the vectors. Result is a scalar.
Cross Product of and :
Result is a vector perpendicular to both and .
TrapDividing by instead of . That gives the scalar projection , which is a number, not a vector. Read which one the question wants.
TrapForgetting the . The cross product's length is the parallelogram, and is sitting right there looking like an answer.
TrapThe minus on the component. Dropping it gives , and the sanity check catches it instantly: must be .
TrapMissing that appears in both vectors, so the last term also carries a . Treating it as a constant gives .
TrapAnswering : , not . is where the *sine* is .
Line through point with direction vector :
Plane through point with normal vector :
or equivalently: , where
Note: Angle between two planes equals the angle between their normal vectors.
TrapSign errors in the direction vector. Notice the and parameters cancel, so the whole equation collapses to ; if yours does not collapse, recheck .
Trap'Parallel to the vectors' does not mean the vectors are normals. They lie *in* the plane; their cross product is the normal.
TrapForgetting to move to the same side. The formula measures , so a plane written as contributes , not .
The official answer is wrongThe commented-out source gives . The numerator is , not , so the distance is . This may be exactly why the problem was commented out.
Commented out on the official page, so it does not appear there. The skill is still on the syllabus, so it is restored here.
TrapAny parametrisation is correct; yours may look nothing like this one. Verify by plugging back into both plane equations rather than comparing to a printed answer.
TrapReading coefficients off a plane written as and forgetting the missing is a coefficient of , not of .
TrapUsing for both lines. Two lines can cross at a point they reach at different times; forcing one parameter finds only the cases where they arrive together, and reports skew for lines that do meet.
A quadric surface is the graph of a second-degree equation in three variables.
| Surface | Standard Equation | Key Features |
|---|---|---|
| Ellipsoid | All traces are ellipses | |
| Elliptic Paraboloid | Opens upward (or downward if negative) | |
| Hyperbolic Paraboloid | Saddle shape | |
| Hyperboloid (1 sheet) | Connected, one piece | |
| Hyperboloid (2 sheets) | Two separate pieces | |
| Elliptic Cone | Double cone, vertex at origin |
Traces are curves of intersection with planes parallel to coordinate planes.
TrapOne sheet vs two sheets is decided by the count of minus signs, not by which variable is isolated. One minus = one sheet (connected); two minuses = two sheets.
TrapMoving the completion constants the wrong way. The and come out on the left, so they get added to the right: .
TrapCalling it a cone. A cone has , not ; the first power is what makes it a paraboloid.
TrapStopping at . A trace is named by its standard form, and the denominators change once you divide through.
TrapReading it as a hyperboloid. Same sign pattern — the right-hand side is the entire difference: one sheet, two sheets, the cone between them.
A vector-valued function (or vector function) has the form:
The graph is a space curve traced out as t varies.
The limit of a vector function is computed component-wise:
TrapThe bracket types. allows equality (closed at ); never does (open at ).
TrapWriting for the first component out of reflex. .
TrapReporting both as holes. is not a hole, it is outside the domain the square root already imposed.
TrapCalling it a helix. A helix needs to move; here .
TrapDeclaring the third component undefined because oscillates. Bounded times something going to zero still goes to zero.
For vector functions and :
If is the position vector at time t:
TrapForgetting that is a vector of three separate constants, not one scalar.
TrapUsing again for the tangent line. The line has its own parameter; reusing makes you solve on the curve instead of on its tangent.
TrapSkipping the normalisation. is a unit vector by definition; alone is not one here.
TrapDifferentiating to . The chain rule puts a minus in front, though here it squares away — the sign matters for direction, not for speed.
TrapThis exact vector is the numerator of the curvature formula — recognise it and Lesson 8 gets shorter.
TrapAssuming an ellipse is traversed at constant speed. It is not — the speed here is , which runs between and . Only the circle has constant speed under this parametrisation.
Arc Length of curve from to :
Curvature measures how quickly the curve changes direction:
For a plane curve :
TrapLosing the Pythagorean identity. ; if your integrand still contains , the collapse was missed.
TrapUsing — that is the circle of radius . Climbing in straightens the curve, so the helix is always less curved than its shadow.
TrapAssuming a curve that visibly bends must have somewhere along it. is *flat to second order* at the origin — it is an inflection point, and curvature is exactly zero there.
TrapIntegrating from to . The interval is , which has length , not .
TrapHunting for a closed form. Most arc lengths do not have one; the exam asks you to set it up, so stop when the integral is right.
A function of two variables assigns to each ordered pair (x, y) in a set D a unique real number .
For :
TrapWriting a strict inequality. The boundary circle is fine — it just makes there. Strictness belongs to logs and denominators.
TrapCalling them circles. The squashes the -direction; equal denominators would be needed for circles.
TrapExpecting closed curves. Level curves of a saddle-ish surface are unbounded; only bowl-shaped surfaces give loops.
TrapIncluding the line. At the log is , undefined — this is the one place the boundary is genuinely out.
TrapA level *surface* of a three-variable function, not a level curve. One dimension up: carves a surface out of space.
We write , if f(x,y) approaches L as (x,y) approaches (a,b) from any direction.
Find two different paths that give different limits:
f is continuous at (a,b) if:
TrapA finite pile of paths never proves a limit exists — but it also never proves it fails. When every path you try agrees, switch to a bound.
The official answer is wrongThe official answer says the limit does not exist. It does: the AM–GM squeeze above bounds the expression by . (Compare , which genuinely fails along — one power of away.)
TrapDegree counting is the fast check: the numerator is degree , the denominator degree . Numerator degree strictly larger the quotient is squeezed to . Only equal degrees can produce a path-dependent limit.
The official answer is wrongThe official answer says the limit does not exist. It does: it is , by the one-line bound . The classic DNE example is — numerator and denominator both degree .
TrapLosing the sign. The term enters as , so the limit is , and would give .
TrapThe substitution is legitimate here precisely because and only ever appear in the combination . It is not legitimate in general.
TrapPolar only settles it when the factor is bounded and the factor goes to zero on its own. If survives in the limit, polar has told you it does not exist.
Partial derivative with respect to x (treat y as constant):
Partial derivative with respect to y (treat x as constant):
If and are both continuous:
TrapGrinding through the term. It has no and no , so it is annihilated by the very first derivative.
TrapEvaluating at before differentiating. for all , so that route reports — you must differentiate symbolically first, then substitute.
TrapDifferentiating the exponent as if were a variable too. In , is a constant — that is the entire content of 'partial'.
TrapBelieving Clairaut always applies. It needs continuity of the mixed partials; the standard counterexample has .
TrapDropping the inner derivative . The factor then has to be cleared of its fractions to reach .
If where and :
If where and :
If defines z implicitly:
TrapSubstituting first is fine here and often faster — but the exam wants the tree diagram, and for a general you cannot substitute at all.
TrapThe minus sign in . It is not optional and it is the single most-dropped symbol in 15.4.
TrapMissing that multiplies two branches ( feeds into both terms), so appears once with coefficient .
Trap is what kills this one. Check the point before doing any algebra — sometimes it collapses the whole expression.
TrapWriting by mistake; that one is , and the factors of are the tell.
TrapSkipping the check. When the surface has a vertical tangent plane there and does not exist — which is exactly what the 'or explain why not' is fishing for.
The directional derivative measures rate of change in direction of unit vector :
TrapDotting with unnormalised, which gives — five times too big. The directional derivative is only a rate if the direction has length .
TrapGiving the maximum rate as a vector or the direction as a number. The rate is ; the direction is (or its unit version).
TrapReporting the rate as negative. 'Rate of decrease' is already signed by the word; the magnitude is the number wanted.
TrapRationalising carelessly: , not .
TrapTreating as a computational accident. A zero directional derivative always means you are moving tangent to the level curve.
Equation of tangent plane to at :
For :
For :
TrapReporting only . The question asks for points on the surface, so the -coordinate has to be computed.
TrapLeaving off . The tangent plane has to pass through the point of tangency; without the constant it passes through the origin instead.
TrapSign of . Moving from to is ; getting that backwards gives , and both look plausible.
TrapWriting for . The differential is the linear estimate; is the true change, and they differ by higher-order terms.
TrapReading and the sign on the last slot. Both and are normal; the form that comes out of is the one asked for.
Point (a,b) is a critical point if:
Let the discriminant be:
Step 1: Find all critical points inside region D
Step 2: Find extreme values on the boundary of D
Step 3: Compare all values - largest is absolute max, smallest is absolute min
TrapDividing by . That silently discards the branch — and here it is a whole critical point.
TrapSquaring is required: . Using unsquared gives here too, by coincidence — the habit will fail you elsewhere.
TrapIgnoring the interior critical point because the boundary already gave a max and a min. The absolute minimum here is *inside*, at .
Trap depends on here, so it must be re-evaluated at each point. A single computed cannot classify two different critical points.
Trap is the one case that is fully decisive with no second test needed. Do not go looking at afterwards; its sign is irrelevant when .
TrapForgetting the four corners. Both extremes here happen to be corner points, which the edge parametrisations only catch at their endpoints.