Scribd
Upload
6 views3 pages

PP 29

Uploaded by

dilipkumar896085
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
6 views3 pages

PP 29

Uploaded by

dilipkumar896085
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

PP 29 : Mixed Partial Derivatives, Mean Value Theorem and Extended Mean Value theorem

The following two definitions are used in this problem sheet.

Definition 1 : Let f : R2 → R. We say that f is convex if f [(1−λ)X +λY ] ≤ (1−λ)f (X)+λf (Y )


for every X, Y ∈ R2 and every 0 ≤ λ ≤ 1. (Geometrically, if we take two points (X, f (X)) and
(Y, f (Y )) on the graph of f , then the graph of f lies below the line segment joining the two
points chosen).
 
a b
Definition 2 : A 2 × 2 matrix M = is said to be non-negative definite if the matrix
c d
 
h
= ah2 + (b + c)hk + dk 2 ≥ 0 for all h, k ∈ R.

multiplication h k A
k
—————————————————————————————————————————
x2 y−y 2 x
1. Let f (x, y) = x+y if (x, y) 6= (0, 0) and f (0, 0) = 0. Show that, at (0, 0),

(a) f is continuous.
(b) fx and fy are continuous.
(c) f is differentiable.
(d) fxy 6= fyx .

2. Let f : R2 → R be differentiable and M ∈ R be such that |fx (X)| ≤ M and |fy (X)| ≤ M
for all X ∈ R2 . Show that |f (X) − f (Y )| ≤ 2M kX − Y k for all X, Y ∈ R2 .

3. (Tangent plane approximation): Let f : R2 → R and (x0 , y0 ) ∈ R2 . Suppose that fx and


fy are continuous and they have continuous partial derivatives on R2 . Let z = L(x, y) be
the equation of the tangent plane for the surface z = f (x, y) at (x0 , y0 , f (x0 , y0 )). Show
that

(a) f (x, y) = L(x, y) + R where R → 0 as (x, y) → (x0 , y0 ).


(b) ey cos x = 1 + y + R where R → 0 as (x, y) → (0, 0).

4. Let f : R2 → R be a differentiable function. Show that f is convex if and only if f (X) ≥


f (X0 ) + f 0 (X0 ) · (X − X0 ) for all X, X0 ∈ R2 (geometrically, the graph of f lies above the
tangent plane at every point on the graph).

5. Let f : R2 → R. Suppose that fx and fy are continuous and they have continuous partial

fxx (X) fxy (X)
derivatives. Then f is convex if, for all X ∈ R2 , the matrix MX =
fyx (X) fyy (X)
is non-negative definite (See the definition given above).

6. Let f : R2 → R and X ∈ R2 . Denote Q(X) = (h2 fxx + 2hkfxy + k 2 fyy )(X). Show that

(a) fxx (X)Q(X) = (hfxx + kfxy )2 (X) + k 2 (fxx fyy − fxy


2 )(X).

(b) fyy (X)Q(X) = (hfyy + kfxy )2 (X) + k 2 (fxx fyy − fxy


2 )(X).

7. Let f : R2 → R. Suppose that fx and fy are continuous and they have continuous partial
derivatives. Show that f is convex if for all (x, y) ∈ R2 the following properties hold

(a) (fxx fyy − fxy )2 (x, y) ≥ 0,


(b) fxx (x, y) ≥ 0 or fyy (x, y) ≥ 0.
8. Show that the function f (x, y) = x2 + y 2 is convex.

9. (*) Suppose that f : R2 → R has continuous second order partial derivatives. For
(x0 , y0 ), (h, k) ∈ R2 , define

H(h, k) = [f (x0 + h, y0 + k) − f (x0 + h, y0 )] − [f (x0 , y0 + k) − f (x0 , y0 )] .

Show that

(a) there exists x between x0 and x0 + h such that H(h, k) = [fx (x, y0 + k) − fx (x, y0 )] h.
(b) there exists y between y0 and y0 + k such that H(h, k) = fxy (x, y)hk.
1
(c) fxy (x0 , y0 ) = lim(h,k)→(0,0) hk H(h, k).
(d) fxy (x0 , y0 ) = fyx (x0 , y0 ).

Practice Problems 29: Hints/Solutions


x−y
1. (a) Note that f (x, y) = 1
+ y1
→ 0 = f (0, 0) as (x, y) → (0, 0).
x

y(x2 +2xy−y 2 ) x(x2 −2xy−y 2 )


(b) If (x, y) 6= (0, 0), then fx (x, y) = (x+y)2
and fy (x, y) = (x+y)2
. At (0, 0),
|y||x+y|2
fx (0, 0) = fy (0, 0) = 0. Now |fx (x, y)| ≤ |x+y|2 = |y| →
0 as (x, y) → (0, 0). This
shows that fx (x, y) → fx (0, 0) as (x, y) → (0, 0). Therefore fx is continuous at (0, 0).
Similarly we show that fy is continuous at (0, 0).
(c) The differentiabilty of f at (0, 0) follows from (b).
fx (0,k)−fx (0,0) −k3
(d) By definition, fxy (0, 0) = limk→0 k = limk→0 k3
= −1. Similarly, verify
that fyx (0, 0) = 1.

2. Follows from the mean value theorem.

3. The equation of the tangent plane is z = L(x, y) where, for any (x, y) ∈ R2 , L(x, y) =
f (x0 , y0 ) + f 0 (x0 , y0 ) · (x − x0 , y − y0 ).

(a) By the EMVT there exists some C lying on the line segment joining (x, y) and (x0 , y0 )
such that f (x, y) = L(x, y) + R(x, y) where R(x, y) = 12 [(x − x0 )2 fxx + 2(x − x0 )(y −
y0 )fxy + (y − y0 )2 fyy ](C). By the continuity of the second order partial derivatives
of f , R(x, y) → 0 as (x, y) → (x0 , y0 ).
(b) Let (x0 , y0 ) = (0, 0) and apply (a).

4. Suppose that f is convex. Let X, X0 ∈ R2 and λ ∈ [0, 1]. Then f (X0 + λ(X − X0 )) ≤
f (X0 )+λ(f (X)−f (X0 )). This implies that λ1 [f (X0 +λ(X −X0 ))−f (X0 )] ≤ f (X)−f (X0 ).
Therefore λ1 [f (X0 + λ(X − X0 )) − f (X0 )] − f 0 (X0 ) · (X − X0 ) ≤ f (X) − f (X0 ) − f 0 (X0 ) ·
(X − X0 ). Allow λ → 0+ .
Conversely, suppose that f (X) ≥ f (X0 ) + f 0 (X0 ) · (X − X0 ) for all X, X0 ∈ R2 . Let
X1 , X2 ∈ R2 and X0 = (1 − λ)X1 + λX2 for some λ ∈ [0, 1]. Then, by the assumption,
f (X1 ) − f (X0 ) ≥ f 0 (X0 ) · (X1 − X0 ) and f (X2 ) − f (X0 ) ≥ f 0 (X0 ) · (X2 − X0 ). From these
two inequalities we get that (1 − λ)f (X1 ) + λf (X2 ) − f (X0 ) ≥ 0. This proves the convexity
of f .

5. This follows from the EMVT and Problem 4.

6. Trivial.
7. Follows from Problem 5 and Problem 6.

8. By applying either Problem 5 or Problem 7 we see that f is convex.

9. (a) Define g(x) = f (x, y0 + k) − f (x, y0 ). Then H(h, k) = g(x0 + h) − g(x0 ). By the
MVT (for one variable), there exists x ∈ R, between x0 and x0 + h, such that
g(x0 + h) − g(x0 ) = g 0 (x)h. Note that g 0 (x) = fx (x, y0 + k) − fx (x, y0 ). This proves
(a).
(b) Again apply the MVT for one variable to obtain (b).
(c) By the continuity of fxy , we have fxy (x0 , y0 ) = lim(h,k)→(0,0) fxy (x0 + h, y0 + k)
= lim(h,k)→(0,0) fxy (x, y). Apply (b).
1
(d) By exchanging the rolls of x and y, we show that fyx (x0 , y0 ) = lim(h,k)→(0,0) hk H(h, k).

You might also like