Scribd
Upload
28 views2 pages

Algorithm Guidelines

Algo

Uploaded by

Kim Taehyung
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
28 views2 pages

Algorithm Guidelines

Algo

Uploaded by

Kim Taehyung
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/ 2

Algorithm Write-up Guidelines

Throughout the class, you will be asked to write-up an algorithm you have
designed to solve a problem. Whether on a homework assignment, project, or an
exam, your solution is expected to contain the following sections. Each section
should be long enough to contain a full, readable explanation of your solution,
and no longer. Remember that the burden of proof is on you. If what you are
explaining isn’t clear to you, you will have no hope of convincing anyone else. On
the other hand, when presenting algorithms, more detail is not necessarily better.
The golden standard is concise but unambiguous.
You may assume that the reader is familiar with the problem. You may make
use of any standard data structures (linked lists, binary trees, heaps, etc), as well
as basic sorting and searching algorithms (linear search, binary search, insertion/s-
election/quick sort, etc) without explaining how to implement them.

• Description. Present a consice and unambiguous description of the algo-


rithm in plain English. You want to explain your general strategy here so
that if you make a mistake in writing it up more precisely in pseudo code
(see below), the grader can ascertain your real intent and give partial credit.
This is NOT the place to restate your pseudo code in English (actually there
is never a place for that). In general, you should not explain what you are
doing, but why you are doing it. Obvious technical details should be kept to
a minimum so that the key computational issues stand out and implemen-
tation details (if any) should be left to the pseudocode section. It might be
useful to include an example to illustrate your approach.

• Pseudocode. A pseudocode description of your algorithm, which does not


rely on a specific programming language’s syntax, but instead uses English
phrases where appropriate. A guiding principal here is to remember that
your description will be read by a human, not a compiler. You should leave
out unnecessary low-level implementation details. For example, ”insert x
at the end of the list” is much clearer than ”list.insertAtEnd(x).” On the
other hand, the pseudocode should give sufficient detail to make the analysis
straightforward (e.g. don’t bury loops in English phrases) while remaining

1
at a high enough level so that it can be easily read and understood. Be
sure that the interpretation of your pseudocode is unambiguous and that
you include explicitly the input and output values.
Here are a few tips:

– Instead of giving a type declaration (e.g., ”double x”), explain a vari-


able’s purpose (e.g., ”x holds the price of the current commodity”).
Type is an implementation detail that’s not necessary here.
– Replace formal control structures
(e.g., ”for (int i = 1; i <= 3*n; i += 3)” with more intuitive ex-
planations (e.g., ”Let i run from 1 up to 3n in steps of 3”)
– Directly employ standard algorithms (e.g., ”Apply quick sort to sort the
node labels in increasing order”), instead of writing out the quick-sort
loop in pseudo code.

While you should avoid unnecessary explanations whenever possible, be sure


to include enough information so that your intent is clear and unambigu-
ous. For example, consider the instruction ”Repeatedly remove the highest
weight edge from G until the graph is no longer connected.” While this is
mathematically well defined, the questions of how to (a) find the highest
weight edge and (b) determine whether the resulting graph is connected are
both nontrivial. These steps would need to be explained in greater detail.
Trust that your readers are competent programmers who can implement the
low-level structures if they understand what needs to happen. Thus the
golden rule for the level of details is - can you implement the algorithm just
from this pesudo-code specification?

• Time Analysis. A worst-case analysis of your algorithm’s running time.


This is where you should explicitly state the data structures that are being
assumed and how their preprocessing and/or query times affect your analysis.

• Proof of Correctness. A proof that your algorithm does what you say it
does and is optimal (if appropriate). If this is a proof by induction, be sure
to state clearly what the base case, inductive hypothesis, and induction steps
are. If this is a proof by contradiction, state clearly what the assumption
to be contradicted is. In general, be sure to define any terms used and
construct a logical argument using complete mathematical sentences. Try
to avoid rambling about obvious or trivial elements and focus on the key
elements. A good proof provides a high-level overview of what the algorithm
does, and then focuses on any tricky elements that may not be obvious.

You might also like