1/15
Looks like no tags are added yet.
How do you take a partial derivative?
Differentiate with respect to the chosen variable while holding all other independent variables fixed. Use the usual product and chain rules for factors that depend on that variable.
For f(x,y)=y exp(x²y), find fₓ and fᵧ.
fₓ=2xy² exp(x²y). For fᵧ, both the prefactor y and the exponent depend on y, so fᵧ=(1+x²y) exp(x²y).
What order does fₓᵧ use in the lecture notation?
Differentiate first with respect to x, then y: fₓᵧ=∂(∂f/∂x)/∂y. Likewise, fᵧₓ differentiates first in y. The mixed partials agree when the second partial derivatives are continuous near the point.
For f(x,y)=x³y², what are the second partial derivatives?
fₓₓ=6xy², fᵧᵧ=2x³ and fₓᵧ=fᵧₓ=6x²y.
What does the gradient tell you about a function and its level surface?
∇f is the vector of first partial derivatives. At a point where f is differentiable and ∇f≠0, ∇f points in the direction of steepest increase and is normal to the level surface through the point.
What is a normal to the sphere x²+y²+z²=1 at (a,b,c)?
Treat the sphere as the level surface f=1, where f=x²+y²+z². A normal is ∇f(a,b,c)=(2a,2b,2c).
How do you differentiate f(x(t),y(t),z(t))?
Use df/dt=fₓx′(t)+fᵧy′(t)+f_z z′(t). Evaluate the partials along the curve. Include one term per intermediate variable.
What is the multivariable chain rule when x and y depend on u and v?
For h(u,v)=f(x(u,v),y(u,v)), hᵤ=fₓxᵤ+fᵧyᵤ and hᵥ=fₓxᵥ+fᵧyᵥ. Evaluate f’s partials at (x(u,v),y(u,v)). When z also depends on u and v, include its term in each sum.
Use the chain rule to differentiate h(t)=f(t²,eᵗ) for f(x,y)=x²y.
Since fₓ=2xy and fᵧ=x², h′=2xy·2t+x²eᵗ. Substituting gives h′=4t³eᵗ+t⁴eᵗ, which also follows by differentiating h=t⁴eᵗ directly.
What is the quadratic approximation to f(x,y) near (a,b)?
Write h=x−a and k=y−b. Use f(a,b)+fₓh+fᵧk+(fₓₓh²+2fₓᵧhk+fᵧᵧk²)/2, evaluating all partials at (a,b). The linear approximation keeps just the constant and first-order terms.
Why is there a factor of 2 on the mixed term in the two-variable Taylor formula?
Apply one-variable Taylor to g(t)=f(a+th,b+tk). In g″(0), the chain rule gives one hk term from differentiating fₓ and one from differentiating fᵧ. For smooth f, the mixed partials agree, giving 2fₓᵧhk. Set t=1.
How do you define a local maximum or minimum for f(x,y)?
At a local maximum (a,b), every sufficiently nearby domain point has f(x,y)≤f(a,b). At a local minimum, every sufficiently nearby domain point has f(x,y)≥f(a,b).
Why must a differentiable interior local extremum satisfy fₓ=fᵧ=0?
Hold y fixed and apply the one-variable stationary-point condition along the x slice. Hold x fixed and repeat along the y slice. A critical point with both partials zero is only a candidate for a local extremum.
How do the slides classify a critical point using second partials?
For smooth f, let A=fₓₓ, B=fₓᵧ and C=fᵧᵧ at the critical point. If B²−AC is negative, a positive A+C gives a local minimum and a negative A+C gives a local maximum. If B²−AC is positive, the point is a saddle. If B²−AC is zero, the test is inconclusive.
What is a saddle point and why does the quadratic test find one?
A saddle has a local maximum along one direction and a local minimum along another. At a critical point the linear terms vanish, so when the quadratic term takes both signs, moving in different directions gives values above and below the value at the point.
Find and classify the critical points of f(x,y)=−x⁴−y⁴+4xy.
Solve y=x³ and x=y³ to get (0,0), (1,1), (−1,−1). Here A=−12x², B=4, C=−12y². At (0,0), B²−AC=16 gives a saddle. At both other points, B²−AC=−128 and A+C=−24 give local maxima.