Home > Archive > Scheme > January 2006 > The spike in the SICP Video Lectures
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
The spike in the SICP Video Lectures
|
|
| bobueland@yahoo.com 2006-01-20, 3:58 am |
| The SICP Video Lecture 4a is called "Pattern Matching, Rule Based
Substitution". For the other lectures there is given a reference to the
SICP book, but not to this one. Also 4a is IMO much more difficult to
understand then the others. So my question is if there is some material
covering the things that Sussman talks about here.
Bob
---
Student: How do you think in order to solve this problem?
Professor: You stare at it untill the solution occurs to you.
| |
|
|
bobueland@yahoo.com wrote:
> The SICP Video Lecture 4a is called "Pattern Matching, Rule Based
> Substitution"... So my question is if there is some material
> covering the things that Sussman talks about here.
I watched it a while ago so I can't remember exactly what's contained
in 4a. I switched over to berkeley's webcasts because for where I am
right now, they are more palatable. Miss the wizard music though.
If you go here:
http://webcast.berkeley.edu/courses...esid=1906978233
the mis-named lecture "Programming paradigms" (12/05) is all about
rule-based substitutions. Might also want to check out "Logic
Programming" (12/7).
| |
| bobueland@yahoo.com 2006-01-21, 6:59 pm |
| No that's not the same thing (I think) because here Sussman talkes
about new kind of deriv-rules looking like
((dd (c? c) (? v)) 0))
and he uses expressions like
dd( (: x1) (: v)). He talkes a lot about a diagram that looks like
Rule
Pattern ----------------> Skeleton
| |
| |
| match | instantiation
| |
| |
v v
Expression --------------> Expression
Source Target
The reason I call this lecture a spike is becuase the difficulty level
for the other lectures was more or less constant, but here it rose to
much higher level, and then went back to the normal difficulty level in
the next lecture. Now I would like to understand this lecture, because
it seems very important, but I couldn't find anything matching this in
the book (and normally a ref is given on the video in the beginning,
but not for this one).
Bob
| |
|
|
bobueland@yahoo.com wrote:
> No that's not the same thing (I think)
Same ideas, different presentation. Rule-based programming, along the
lines of regular expressions.
Brian Harvey's version:
1. doesn't require calculus knowledge, and
2. uses more human words in making queries
3. defines rules that involve family relationships and food
Gerald Sussman's version:
1. requires Calculus knowledge
1. defines rules for getting a derivative
2. uses an arguably more practical example of a query language, albeit
a much more advanced one.
>because here Sussman talkes
> about new kind of deriv-rules looking like
> ((dd (c? c) (? v)) 0))
Brian Harvey uses this same question mark notation.
> and he uses expressions like
>
> dd( (: x1) (: v)).
If memory serves, Sussman defined the colon to be a substitution
object.
>He talkes a lot about a diagram that looks like
>
> Rule
> Pattern ----------------> Skeleton
> | |
> | |
> | match | instantiation
> | |
> | |
> v v
> Expression --------------> Expression
> Source Target
Having used regular expressions pretty intensively for a former job, I
find diagrams like this needlessly complicated. HOWEVER, I'm sure that
other people find the diagram clarifies things.
When I learned to talk, I didn't learn about what the imperative form
was vs. the subjunctive form, and first person vs. third person, and
past participle vs. whatever. My parents didn't say to me "H., today
you're doing to learn about the past imperfect. But first let's
understand exactly what the past imperfect is vs. the past preterite,
in case you learn another language and that becomes an issue." Nope,
they just talked and I figured it out which amazingly is how is works
for most everyone.
And yet everything got more cumbersome and more formal in high school
and college where I learned Spanish, because instead of just having
natural conversations in class and figuring out organically when to use
what, everything was artifically contrived and involved making charts
of every possible different verb form, understanding which verb form
was which, and dissecting things in very formalized ways to arrive at
which word to use. I don't like this method, although a lot of people
apparently do; otherwise it wouldn't be such a popular pedagogical
method
I played around with regular expressions for a bit at a previous time
in my life, glancing over a few pages in a book, but more through
experimentation, gradually getting a feel for them and creating more
complicated expressions, culminating with creating them regularly for a
job in which pattern-matching got pretty knarly.
In other words, it's possible to understand the concepts involved
without understanding this particular chart, which I'm sure had a lot
of merits as well. But if it's just the chart that's throwing you off,
try Brian Harvey's "rule-based substitution" lecture instead. There's
also some SICP lectures that were given at ADUni.org which may cover
pattern matching rule-based subsitution somewhere, although I'm not
sure where:
http://www.aduni.org/courses/sicp/index.php?view=cw
| |
| Brian Harvey 2006-01-22, 3:57 am |
| "H." <hbe123@gmail.com> writes:
>Brian Harvey's version:
Thanks for the compliments, but I want to disclaim authorship; my lectures
on this topic are closely based on SICP 4.4, so they are much more Sussman's
(and Abelson's) than mine! I'll accept responsibility for the part about
food. :-)
| |
| Nils O. Selåsdal 2006-01-22, 7:56 am |
| H. wrote:
> bobueland@yahoo.com wrote:
>
> I watched it a while ago so I can't remember exactly what's contained
> in 4a. I switched over to berkeley's webcasts because for where I am
> right now, they are more palatable. Miss the wizard music though.
>
> If you go here:
> http://webcast.berkeley.edu/courses...esid=1906978233
You happen to know if these are available in some nicer (preferrably
easier downloadable) formats ?
| |
|
|
> You happen to know if these are available in some nicer (preferrably
> easier downloadable) formats ?
Unfortunately Berkeley only streams their videos, at least that I've
seen.
But what is the difficulty you have?
| |
| Nils O. Selåsdal 2006-01-22, 10:00 pm |
| H. wrote:
>
>
> Unfortunately Berkeley only streams their videos, at least that I've
> seen.
>
> But what is the difficulty you have?
Off-topic issues, like RealPlayer crashes on my Linux box.
mplayer won't play it, lost quicly interrest tracking the dependencies
needed.
I've yet to figure out what's needed for vlc to play it, and xine won't
compile(newest gcc barfs badly on it).
| |
|
| >> Off-topic issues, like RealPlayer crashes on my Linux box.
I'm curious now if this is your setup or a known Linux / RealPlayer
issue. If the latter, then I think it should really behoove someone who
works on Berkeley webcasts to find an acceptable solution!
| |
| Kjetil S. Matheussen 2006-01-22, 10:00 pm |
| | |
| Nils O. Selåsdal 2006-01-24, 7:57 am |
| H. wrote:
>
> I'm curious now if this is your setup or a known Linux / RealPlayer
> issue. If the latter, then I think it should really behoove someone who
> works on Berkeley webcasts to find an acceptable solution!
Got it going - needed to turn off Xvideo in the RealPlayer.
|
|
|
|
|