Instructions & Guidelines on Doing and Submitting Homework.
==========================================================
(0) Homeworks must be submitted as hard-copy. They can be submitted to me
(personally in or after class, or under my office door, or in my Math
Department mailbox) before 5 pm on the listed Due-date. If you are
off-campus and cannot give me a hard-copy that afternoon, then you can
submit the HW electronically via email to evs@math.umd.edu for the
time-stamp, but you must hand the hard-copy to me at the next class.
(1) You are expected to solve the homework problems essentially by yourself,
although you may give each other hints and pointers. (As a typical example,
you may certainly ask or answer a question like "Is there an R command or
trick to do _______ ?") You may also get hints or ideas from me or other R
users around the department or campus, and you may of course online chat-lines
and R user groups. Many command requests can simpled be Googled. BUT you are
expected to write up the submitted (hard-copy) homeworks COMPLETELY by yourself.
(2) A good way to work in R generally is to copy the successful R statements
and the output they generate into a text-editor (eg, vi or emacs) log-file.
But please do not hand in huge log-files and especially do not hand in the
listings of very large R objects. The guideline is this: hand in exactly as
many R code and output lines as needed to show how you generated the result
AND NO MORE. This involves editing the log of your R session to include the
lines of R code which can efficiently perform the requested tasks, and then
either a subset or summary of the R objects you generate.
(3) When large R objects are generated, e.g. where you generate R objects
consisting of hundreds of lines each, DO NOT LIST ALL OF THE OUTPUT. But it
would be a good idea to show some parts (and/or dimension, and/or summary) of
each requested R object.
(4) It is up to you to edit the log-files to show the results clearly and
succinctly. INTELLIGENT PRESENTATION IS PART OF THE HOMEWORK TASK (at least 10%).
(5) Especially as we progress to more sophisticated R tasks, a very important
part of your work is to show that your computed results are correct. So if you
can think of an auxiliary computation or intermediate result to display in
some way, to show that the computation is working as it is supposed to, then
do it and include it ! As an example: if you have computed a vector in two
different ways, x1 and x2, then you can prove they are the same in a very short
space by displaying that the value of sum(abs(x1-x2)) is 0 to a suitable
accuracy (usually of order 1.e-10 or less).
I will reserve 10-20% of the credit on each Homework to this checking: that is,
you will lose this much credit if you do not find some way of showing
(without listing everything out) that your computed result is the correct one.