next up previous
Next: About this document

Name:

Directions: Show your work! Answers without justification will likely result in few points. Your written work also allows me the option of giving you partial credit in the event of an incorrect final answer (but good reasoning). Indicate clearly your answer to each problem (e.g., put a box around it). Good luck!

Problem 1 (25 pts). Determine whether the following sequences and series are convergent or divergent (give valid reasons!). For those which are convergent find the limit.

  1. displaymath464

    We can rewrite this as

    displaymath465

    Convergent, by exhibition of the limit!

  2. displaymath466

    displaymath467

    Convergent, as a monotone and bounded sequence converges to a limit.

  3. displaymath468

    For this infinite series to converge, we must have that the terms tend to zero as tex2html_wrap_inline572 : i.e.,

    displaymath469

    This is clearly not the case, as the argument to cosine would have to approach multiples of tex2html_wrap_inline574 , and an integer argument of n will not do so! Therefore, the series is divergent.

  4. displaymath470

    This simplifies to

    displaymath471

    and we can show that

    displaymath472

    since tex2html_wrap_inline578 is a continuous function of its continuous argument on this interval tex2html_wrap_inline580 . Hence, the sequence is convergent with limit 0.

  5. displaymath473

    This is divergent, by comparison with

    displaymath474

    displaymath475

    And since tex2html_wrap_inline582 diverges, so does the given series.

    You could also proceed by the limit comparison test with tex2html_wrap_inline584 , and then show that

    displaymath476

    diverges by comparison with the harmonic series, for example.

Problem 2 (5 pts). Express tex2html_wrap_inline586 as an infinite series.

displaymath477

displaymath478

As a check, evaluate

displaymath479

Problem 3 (10 pts). Analyze the integral

displaymath480

determining conditions on p for which it converges.

You should use the definition of the improper integral to do your analysis (that is, use limits).

First of all, we note that if tex2html_wrap_inline590 then the integral is perfectly well behaved. Hence we consider in detail only the cases p>0.

  1. p=1 is special case:

    displaymath481

    which diverges, since the limit does not exist as tex2html_wrap_inline596 of tex2html_wrap_inline598 .

  2. p<1:

    displaymath482

    converges.

  3. p>1:

    displaymath483

    diverges, since the anti-derivative blows up as tex2html_wrap_inline596 .

Thus, the answer is that for p<1 the integral is perfectly convergent; otherwise it diverges.

Problem 4 (10 pts). Give examples of series for which

displaymath484

displaymath485

displaymath486

displaymath487

  1. displaymath488

  2. displaymath489

  3. displaymath490

  4. displaymath491

Problem 5 (10 pts).

  1. State in precise terms (i.e., using tex2html_wrap_inline608 ) what it means to say that the sequence tex2html_wrap_inline610 converges to the limit L.

    displaymath492

  2. From your definition (and not using the algebra of limits) prove that

    displaymath493

    Given tex2html_wrap_inline614 .

    Consider

    displaymath494

    We now find where

    displaymath495

    We take N to be the first positive integer greater than x. Then

    displaymath496

    Q.E.D.

Problem 6 (10 pts). Decide whether the following integrals are convergent or divergent (give reasons!):

  1. displaymath497

    displaymath498

    The problem occurs at x=2, where the integral is improper. I should be written as the sum of two integrals,

    displaymath499

    Neither one converges, as we can see by comparison, for example:

    displaymath500

    which, upon u-substitution, looks like

    displaymath501

    (and which fails to converge).

  2. displaymath502

    This converges by comparison:

    displaymath503

Problem 7 (5 pts). Find the values of p for which the series is convergent:

displaymath504

Obviously if tex2html_wrap_inline628 we're divergent, as the terms will grow at least linearly. Hence consider p<0. We'll use the integral comparison test. Consider

displaymath505

Now f is clearly positive on the interval tex2html_wrap_inline580 , but it needs to be eventually decreasing if we're to use this test. Thus we consider the derivative:

displaymath506

This must be negative for large values of x: that is, in the limit as tex2html_wrap_inline638 , tex2html_wrap_inline640 . Thus

displaymath507

Thus, only in the event that tex2html_wrap_inline642 will this condition be satisfied. For tex2html_wrap_inline644 , f is increasing, which means the that series diverges (since the limit of the terms cannot be zero). Thus, we can now restrict to tex2html_wrap_inline642 .

displaymath508

We use a u-substitution tex2html_wrap_inline652 to write this as

displaymath509

which converges if and only if p<-1. This then, is our answer: the series converges if and only if p<-1.

Problem 8 (15 pts). Consider the series

  equation427

  1. Demonstrate that the series converges, and use your calculator to estimate the limit.

    This converges by the alternating series test, as

    1. displaymath510

      displaymath511

      and

      displaymath512

    2. The calculator will give rather strange results, however, as this series converges slowly: I considered tex2html_wrap_inline658 and tex2html_wrap_inline660 , and got an average of .601481. The partial sum tex2html_wrap_inline664 !

  2. Plot the first 10 terms of the sequence tex2html_wrap_inline666 and the first 10 partial sums of the series (1) on the same axes. Indicate clearly which graph is which.

       figure321
    Figure 1: Problem 8: Terms in black, sums in blue.

  3. How many terms m of the series (1) do we need to add in order to assure that the size of the error incurred by approximating the series by the tex2html_wrap_inline670 partial sum is less than .0001?

    In order to assure that the error is less than .0001, we need to find m such that

    displaymath513

    Hence we solve tex2html_wrap_inline676 , or

    displaymath514

    Thus, tex2html_wrap_inline678 terms should work.

Problem 9 (10 pts). Consider the series

  equation441

  1. Use the sum of the first 20 terms to estimate the sum of the series.

    1.2008678419584369

  2. Bound the error of this estimate (that is, find values a and b such that tex2html_wrap_inline684 ``tightly'').

    displaymath515

    displaymath516

    Hence

    displaymath517

Problem tex2html_wrap_inline686 (Extra Credit: 5 pts). There are 2 million points in the plane, no three of which are colinear. Is it possible to draw a line through them so that exactly 1 million are on one side, and 1 million are on the other?




next up previous
Next: About this document

LONG ANDREW E
Tue Nov 21 16:44:21 EST 2000