CS L320
Fall, 2008
hw1
Parts 1, 2 and 3 are due Thursday, September 4. Parts A, B, C, ... are
Tuesday, September 9.
Written work must be
-
legible - handwritten papers are OK since writing
mathematics in most word processing systems is a painful distraction
from the ideas.
-
logically clear (of course) - the point of the arguments you make is
not to convince me (I already know) or even to convince yourself (it
may be obvious to you) but to convince me that you have convinced
yourself for good reasons.
-
free of spelling and grammatical errors - I will make
some allowance made for non
native speakers of English, but I expect your English to improve as
the semester progresses.
-
honest - if you are confused at some point, say as clearly as you can
what you are confused about. Don't guess - a correct guess without a
good reason may convince you that you understand the problem when that is
not in fact true. An honest description of what puzzles you will teach
you more, and let me teach you more.
-
Apply for a cs320 account on the CS Department's unix system. Apply
for that account even if you already have a unix account, so that you
are put on the proper mailing lists.
I will use your cs email address for all electronic communication,
so be sure to check it regularly, or forward from it. If you write me
from that address I will surely get your email. If you write me from
some other address your mail may end up in my spam bucket.
-
Answer the question I've asked about the sudoku
puzzle from the Globe.
-
Fill out the questionnaire on
the course web page at
www.cs.umb.edu/~eb/320/questionnaire.html
Some problems from Bender &
Williamson. Which ones depend on how far I get. Specification is B&W
[pages]/[problems].
You should read the book if you find that it helps you understand the
material and do the problems. You may find that you don't need to.
- B&W Lo-10,11/11,12,14,16,23 (for 12 do the
converse only, not the inverse).
- B&W SF-12/5
- B&W SF-14/13. (If you can, prove (e) two ways, first directly
using some method of your choice, then as a consequence of (a)-(d).
-
In class we discussed representing subsets of a (finite) set as bit
vectors. How might this representation provide another way to attack
the previous problem.
(If you've had abstract algebra you should
be able to answer this in one line. If not, try to find an answer
anyway.)
- If you ask n people a yes/no question, there are |P([n])| =
2n ways the answers could turn out - each person could
answer "yes" or "no". Another way to say that is that
there are 2n ways to divide an n element set in
two pieces, if the pieces are to be labelled and empty pieces are
allowed - the "yes" piece and the "no" piece.
Suppose the question could be answered yes/no/maybe. Then how many
ways are there. That is, how many ways are there to divide a set with
n elements into three labelled subsets, empties allowed?
- Counting the number of ways to divide a set with
n elements into k unlabelled pieces with no empty pieces
allowed is a
very hard problem. But you can do a few computations. Work out the
answer for n = 3, 4 and 5, k = 1, 2 and 3.
I suggest you do this with a classmate, or at least check your answer
with a classmate.
If you work out a few more examples you might be able find out some
cool stuff from
The On-Line Encyclopedia of Integer Sequences
From: Ethan Bolker
To: "Henry Z. Lo"
CC: cs320-1@cs.umb.edu, "Shane C Woods" ,
"Julia Strangie-Brown" ,
"Johnston, Mary Jane" , eb@cs.umb.edu
Subject: HW1
Date: Mon, 8 Sep 2008 08:31:32 -0400 (EDT)
> From: "Henry Z. Lo"
> To: eb@cs.umb.edu
> Subject: HW1
> Date: Sun, 7 Sep 2008 21:58:49 -0400 (EDT)
>
> Hi Professor,
>
> On part D, do you mean using bit vector representations of subsets
> in order to solve Part C of the homework? I can't seem to think of
> how I can prove equality using bit vectors.
Yes, that's what I mean. Here's a hint about how bit vectors might
help: think about how to compute the bit vector representation of
A +O B from the bit vector representations of A and B.
(A +0 B is the best I can do in ascii to get the + inside the circle.)
> On part E of the HW1, what does it mean if the pieces of the set
> are labeled or unlabeled?
They're labelled - the problem says so.
> Thanks,
> Henry