Mailman 3 May 2001 - Edu-sig

Python Briefing for Faculty
by Seabrook, Richard 12 May '01

12 May '01

I'm putting together a 2-3 hour briefing + demonstration for faculty to try to get them interested in Python for our introductory programming course, which uses a baffling subset of C++ at the moment. I thought I'd do it in a PC lab with Win2000 and use Python 2.0 + IDLE. I teach Python under Linux, but I think the windoze environment will be more convincing for them, and engaging for beginning programming students. I plan to discuss programming environments, provide an overview of the language, and have them type in some small programs. Anyone done this recently? Have any thoughts to contribute? Thanks, Dick S. - rhseabrook(a)mail.aacc.cc.md.us Anne Arundel Community College (410)541-2424 *** Speed the net! http://enterprise.aacc.cc.md.us/~rhs

6 13

Graphics for beginners, and other resources
by Richard Crook 11 May '01

11 May '01

I don't normally shout much about our resources, but the list seems recently to have been discussing areas where we might be able to save folks some work. LiveWires is a computing holiday in the UK for 12-15 year olds; all our course material and Python modules is on our website. In particular, we have a support library to provide (Tk-based) graphics in a window with a straight procedural interface for beginners (similar to that described recently by Daniel Yoo). Daniel: apologies, I haven't yet had time to take a detailed look at what you've done, but you might find our course materials relevant anyway... In addition, some of what Jason Cunliffe posted struck a chord, so there might be some synergy there -- we've been actively using these materials for a couple of years to teach real beginners. I know Jeff Elkner has also picked up on this for some of his students (Hi Jeff!). Website at http://www.livewires.org.uk/python/index.html, for anyone interested. Feel free to comment -- we're in the process of updating and extending for this year's holiday. If anyone's interested in the directions we're going from here, drop me a message. Regards, Richard -- Richard Crook // SJ Consulting Ltd, Cambridge // richard(a)sj.co.uk

2 1

Re: [Edu-sig] Python Briefing for Faculty
by Markus Gritsch 11 May '01

11 May '01

> Thanks for the suggestion! I got a copy of Scintilla some time ago but had > forgotten about it since I do most of my Python stuff under Linux using > vi. Maybe you should get a recent copy of SciTE before evaluating it since lots of things have been improved. Markus

1 0

re: Python Briefing for Faculty
by Arthinator＠aol.com 10 May '01

10 May '01

Kirby writes >In IDLE, I can assign variables and they persist, as do referencesto whatever >imported stuff. As I do this and that interactively, I make use of these references >-- I'm sitting in a pile of tools and they all stay defined from one command to the >next (plus Icreate new ones at will). I am and am not following you. Certainly I agree with you as to the importance of interactive mode, especially when configured for learning. Probably what attracted me most about Python initially was the clean and accessible interactive mode, where I could experiment and get my bearings. But my own work habits now are a text editor I had gotten accustomed to pre-IDLE, and IDLE for interactive mode - often open simultaneously. Never really gave IDLE a chance as my primary editor. Mostly just inertia . But I do see in it the design elegance I admire in Python, and love that its all just Python being Python. Based on my reading good code to learn coding theory, had spent some time with the IDLE code some time back trying to follow it best I could. Actually got a lot out of it. Biggest challenge at that stage for me was sorting out what was Tkinter specific and what was general Python. Have not used scintilla itself, but a quick look at the page indicates it to be a programmers editor. So I wouldn't expect it to have anything to do with interactive mode Python. Which of course is always available to you through IDLE, even with scintilla, etc. open and in use. Anyway, think you should see my point as to why I don't fully see what you are driving at. ART

2 1

Re: Intro and question: assignments/projects for year end
by Brent Burley 10 May '01

10 May '01

HotToPot(a)aol.com wrote: > If its Game of Life, start them with functioning code and challenge > them to extend it in some fairly specific way. > ... > But don't ask them to face the blank page. I've prepared something along these lines. I've written a functioning Reversi game that provides a framework for defining new computer player strategies. It's meant as a general programming project and an introduction to AI. There's an accompanying tutorial. See http://erburley.home.netcom.com/Reversi-index.html Reversi is one of the first games I programmed (in high school, in BASIC on an Apple II+). I rewrote this game several times in various languages on various computers as I learned more about programming and AI. Most high school students would probably rather learn to program arcade games and graphics. I think that's fine and I'm interested in making some materials that could be used to that end. However, for the right student, I think AI programming can be a more interesting programming problem since it causes you to think about how humans think, how computers think, how they are related and how they are different. Brent

1 0

About myself
by Roman Suzi 10 May '01

10 May '01

Hello! Well. I was asked to introduce myself and here is my informal self-introduction in the aspect of this SIG. As I hope to stay (at least lurk) here for long. Two years ago I conducted a class in a 10-11 forms (in Russia, these can be called "higher school" forms). And for no apparent reason I chose Python to do programming. (At that time I had no exposure to Python. I even thought, it was a kind of LISP or Schema or similarly AIsh.) I used Perl to do admin/text processing task, C for other small things. I learned Python for the purpose of teaching only, but it occured to be so magically powerful and attractive, that now (I changed jobs since then) it my primary language for most tasks. However, I was always interested in the teaching programming and it always intrigues me how one become a programmer. Yesterday you can't program, and today you feel, that you can! I can't tell that those programming classes were full success, but nevertheless, I made some observation of what young people expect from the CS/programming teacher. And I must tell you that I heard a lot of things like: "Why not {Delphi, Java, Pascal, C++}" from my pupils. Those usually did not understand what programming is about. On the contrary, those who listened made more progress. They were open to new ideas (not just their maximalist views) and were very supportive of what we did. * Really, I don't know why I am here. Probably, I want to listen to more stories (cases) and find something similar or dissimilar. But one thing I am very strong about: if we need to teach somebody programming, the Python is the choice No. 1. (Well, probably it depends on the age. For little-ones Logo is probably the choice or some special adopted environment). But for those who want to learn programming Python is very-very useful. (It's hard to switch from it to other languages, though ;-) But in schools we do not need to prepare programmers. It's all about development in the minds... OK. I joined this list as I feel that this direction could be forgotten, SIG closed, etc. IMHO, Python's one of primary goals is to be excellent for teaching programming. In today's school we are drifting more and more toward "user skills" (worsprocessing and such mouse-moving behaviour). This is what bothers me too, as I think those who is capable of programming need to have tools to learn and get successful results. This raises their IQ ;-) and improves understanding of life, whereever they will work. Still I do not see what can be done to the situation except with promoting Python to the advanced users and teachers. Unfortunately, I am no more in position where I can influence this process here (I worked in the teacher's retraining institute's IT center), so probably my presence here will not be too visible (I do not have much time to do something real, usually I just generate lots of ideas ;-) I read some of the archives. Probably I could add something later to share with others. (I am also a participant of Seul/EDU (Simple End User Linux, educational aspect. Python Edu-SIG /\ Seul-EDU != []. And a moderator of relcom.education newsgroup. So, if there are some need to communicate some ideas/offers/questions/queries with the later, you can ask me. I will try to translate it into Russian and put it into there. In fact, in the near future I want to prepare an "advertizement" of Python (and free software) for relcom.education. That is why I also want to listen here. Wow. It was quite lenthy explanation! (I do not have a web-page, at least updated less than 4 years ago ;-) Sincerely yours, Roman Suzi -- _/ Russia _/ Karelia _/ Petrozavodsk _/ rnd(a)onego.ru _/ _/ Friday, May 04, 2001 _/ Powered by Linux RedHat 6.2 _/ _/ "A program without bugs is obsolete." _/

3 5

Intro and question: assignments/projects for year end
by HotToPot＠aol.com 09 May '01

09 May '01

>If they only have about two weeks to "learn" Python, and write a small program >or two, what types of things could I have them do? We won't have time to get >into GUIs. These can't be very ambitious projects, due to the limited amount of >time. A suggestion\idea made without great conviction, and probably too general to be of great use: Don't ask them to face a blank page. If its Game of Life, start them with functioning code and challenge them to extend it in some fairly specific way. If its graphics, start with functioning code that draws lines, ask them to draw boxes and circles. Even at high levels of programming, few people are asked to face the blank page. One is extending, optimizing, debugging, porting, etc. Give them something broken to fix. Plant a few bugs. But don't ask them to face the blank page. ART

3 2

Re: Determinants and permutations
by Daniel Ajoy 09 May '01

09 May '01

On 9 May 01, at 12:01, edu-sig-request(a)python.org wrote: > > Now to turn all of that into code. Now I know how I'm > > teaching recursion > > this summer. > > However, if you want to be brute force about it, and non-recursive, > you just need the n! permutations of column elements, picking one > element from each column as you go down the rows. Each selection > nets you a product, and you still need to compute its sign (talked > about that just a post or two back). > > I was relieved to find this, as I was looking at the selective > (row,column) blocking with recursion, and not liking the looks of > that challenge. I did, sometime ago, programmed the recursive definition of determinant using adjoints and minors, I did it in Logo (MSWLogo to be precise), but it was terribly slow. That is really "brute force" approach. I thank Kirby for showing me the permutations approach. Now, that with his help I was able to figure out the pattern in the signs, my determinant implementation in Logo works wonderfully (I tested with a 6x6 matrix). > How do I get the permutations? Well, I'm sure there's lots > of ways, but I just set up a base-4 number (for a 4x4 matrix) > and count from 0000, 0001, 0002 -- except I eliminate any > string with repeating digits, so 0123 is the first admitted. > Add 1 to that, base 4, and you get 0130, which isn't "legal" > either, but you get the idea (add up through 3210, crossing > out any with repeating digits, and you've got your 4!=24 > legal permutations). I could refine this some (i.e. start > with 0123..(n-1) -- but it works well enough for my purposes > as is, so further optimizing is back burner for now. That is the advantage of an object oriented languaje: a lot of the code can be reused easily. I developed my permutations algorithm using the backtracking approach, and its more general. Since I had that algorithm ready, it was fairly easy to develop the determinant algorithm too. My permutations algorithm is called permconj since "conjunto" is "set" in spanish. show permconj [1 [a b c]] [[a] [b] [c]] show permconj [2 [a b c]] [[a b] [a c] [b c] [b a] [c a] [c b]] show permconj [3 [a b c]] [[a b c] [a c b] [b c a] [b a c] [c a b] [c b a]] show permconj [0 [a b c]] [[]] show count [a b c] 3 show count permconj [6 [a b c d e f]] 720 I also have "permu" that returns only the counts. es permu [3 1] 3 es permu [3 0] 1 es permu [3 2] 6 es permu [3 3] 6 Anyway, one of these days I'm going to tray to do Kirbys polyhedra in MSWLogo too, since MSWLogo has 3D graphic rendition incorporated. Daniel

2 1

graphics.py: simplified graphics for beginners
by Daniel Yoo 09 May '01

09 May '01

Hiya everyone, Here's an updated copy of the small graphics module I cooked up, with a small circle-drawing demonstration that uses it. What do people think? Would this module make graphics programming more accessible to beginning students?

1 0

re: [Edu-sig] Algebra + Python
by Lloyd Hugh Allen 09 May '01

09 May '01

In response to the message """ Message: 1 From: "Daniel Ajoy" <dajoy(a)softhome.net> To: edu-sig(a)python.org Date: Mon, 7 May 2001 18:14:44 -0500 Subject: RE: [Edu-sig] Algebra + Python On 7 May 01, at 12:01, edu-sig-request(a)python.org wrote: > A simple way to compute the determinant is to add the signed > products of all permutations of matrix elements, choosing one > from each row (a total of row! possibilities). ciphers.py is > in charge of returning a list of all those possibilities Does that work for matrices 4x4 ? I ask because it didn't work for me. """ [very long. sorry. this is why we iterate when we program, so that the computer can do the boring stuff for us.] There are few "tricks" for finding determinants > 3x3. You have to go back to the why of how the tricks work. Let's start with a 1x1: [a] The determinant of that matrix is <a>. It retains whatever sign the element had had. ===== Easy enough. Go on to a 2x2: a b c d which can be represented as [ [ a , b] [ c , d ] ], if you're a programmer type person and prefer your data one-dimensional. Start in the top row. Put your finger on "a", and block out the row (that is, the element "b") and the column (that is, the element "c"). You are left with the trivial submatrix [d], which has determinant of <d>. This gives us the first part of the determinant, which is a*d (the thing that we had our finger on, times the determinant of the corresponding submatrix). Now put your finger on the next element in the top row, b. Ignore everything else in its row and in its column. You are left with the trivial submatrix [c], which has determinant of [c]. This gives us the second part of the determinant, b*c. Go back through your list of parts of the determinant, and alternate the signs on them. In this case, we have a*d - b*c. ===== Next, for a 3x3: a b c d e f g h i Start in the top row. Put your finger on "a", and block out the row and the column of "a". You are left with submatrix [ [ e , f] [ h , i ] ]. The first part of the determinant will be <a> * det ( [ [ e , f ] [ h , i ] ] ), which is a ( ei - hf ) #taking the 2x2 determinant of the submatrix; the sign of a is #unchanged because a is in column 0. aei - ahf #using the distributive property Now go to "b". Blocking out the row and column of "b" leaves the submatrix [ [ d , f ] [ g , i ] ]. Because this is the second element we have played with, we need to change the sign of b, and then we will multiply that by the determinant of the submatrix. -b ( di - gf ) #the sign of b is opposite because b is in column 1 -bdi + bgf #using the distributive property Similarly, we get the elements c * det [[d,e][g,h]] c ( dh - eg ) cdh - ceg When you string those together, you get the familiar "diagonal" cheating method of finding the det of a 3x3. The procedure above, though, generalizes to n x n. Sorry to be verbose. Determinants are cool because the determinant of a 2x2 gives the area of a parallelogram with one vertex at the origin and two points identified by the rows or columns; or of a 3x3, the volume of a solid with parallel faces with vertices identified similarly; or of a 4x4, the hypervolume... and because if the determinant is zero, you know better than to look for an inverse (which might otherwise be time-consuming). (not that finding the determinant of a large matrix is terribly fast). Now to turn all of that into code. Now I know how I'm teaching recursion this summer.

2 1