Mailman 3 - Edu-sig - python.org

Fwd: Data Structure and Algorithms in Python: May 29: After Class Review
by kirby urner 30 May '22

30 May '22

I share these emails to my students from time to time, to document the progress of an 8th grade online class. The Notebook I was working on today. https://github.com/4dsolutions/elite_school/blob/master/TheCabal.ipynb -- Kirby ========== From: Sunshine Elite Education Date: Sun, May 29, 2022 at 7:39 PM Subject: Data Structure and Algorithms in Python: May 29: After Class Review Greetings all -- Tonight's talk was about how those intricate little patches of symbols, in math books, are pretty much algorithms, but in a pre-computer language. We use letters for numbers because the recipes are for a whole variety of inputs. The gears turn, and we get our computed outputs. That's been going on since long before electronic computers. My opening example involved computing the volume of a tetrahedron, regular or irregular, on the basis of its edge lengths only, of which there are six. I showed the Cayley-Menger algorithm for computing that, which involves taking the "determinant of a matrix" a topic in some high school math curricula (e.g IB, AP) but not all. Python's sympy was able to take that matrix and convert it to a lot of terms under a radical sign (2nd root symbol). Using plane nets, I was able to compute the volume of: * a regular tetrahedron of edges 1 * an "A module" (1/24th of the above) * an "E module" (1/120th of a rhombic triacontahedron) This is shoptalk from a niche geometry I specialize in studying. I was demonstrating how I make use of the same tools I'm teaching, when doing my own geometry research. Python and mathematics weave together in Jupyter Notebooks to give you symbolic operations with sympy, plotting capability (graphing) with matplotlib, and tabular data (pandas -- coming up). During the 2nd half of the talk, we dove deep into Python's capabilities, reviewing the role of __special_names__ ("__ribs__") via the Ornery class <https://replit.com/@kurner/Ornery-Type>, and then diving into @properties, which use @decorator syntax. That magic circle source code we looked at is here: https://github.com/4dsolutions/elite_school/blob/master/magic_circle_v2.py Kirby *Course Name*: Data Structure and Algorithms in Python *Datetime*: 01/29-06/19 Wednesday,Sunday 6:00pm-7:00pm *Location*: Online10 Thanks, Sunshine Elite Education

1 0

Python Certificate
by gabor.guta＠proinbox.com 21 May '22

21 May '22

I have checked out the the Python certification programs and I think it would be interesting to create something more "official" and more open/cheaper. I have found a Polish company which calls itself as Python-Institute (which is a bit misleading name in my opinion as it suggest that it is officially related to some organization of the Python community) and sells tests/certs from approx. fifty to several hundred EUR depending on the level. I have tried the entry level and for me it was disappointing from the perspective of the quality of the questions. I haven't tried Microsoft's certification yet. I think it would be a very viable option to create a certification program which is officially supported by the PSF and provide some reasonable pricing (i.e. free if someone just want do a self assessment and a reasonable price if someone want to take the test in a controlled environment). The scope could be the core language elements. Let me know what you think about the usefulness of the idea.

11 25

documenting the evolution of a 21st century high school course (Silicon Forest)
by kirby urner 16 May '22

16 May '22

An email to students, after Zoom class this evening: ---------- Forwarded message --------- From: Sunshine Elite Education Date: Sun, May 15, 2022 at 9:32 PM Subject: Data Structure and Algorithms in Python: May 15: After class (re Quicksort) To: <kirby.urner(a)gmail.com> Thanks for joining me this evening. The famous QuickSort, which I coded live, is here: https://nbviewer.org/github/4dsolutions/elite_school/blob/master/ADS_sandbo… (scroll down) I was somewhat surprised I could get it working without notes. Note: if the goal really is super speed, then C/C++ is a better bet (I showed C++ for Kids, a book). But Python, with its clear high level syntax, is good for communicating the gist of the algorithm. Python itself is written in C (often) so it's internal, native sorting capabilities (e.g. part of L.sort and sorted(L)) are much faster than those like the one I shared, written in Python. In a practical sense, you don't really need to write a sort algorithm in Python as these are already provided to you. The source code for Python is out there, if you get curious about how it sorts stuff internally. Check into how the list type is implemented -- that would be where I'd poke around (I haven't yet). There should be a book out there that's like a guide to Python's C internals. Is there? I don't know. I did some overview of our evolving online textbook, which now has a TOC. Feel free to explore any time, by way of review. In the Graph Theory sandbox, I've added more links to student work. https://github.com/4dsolutions/elite_school/blob/master/ADS_TOC.ipynb Kirby *Course Name*: Data Structure and Algorithms in Python *Datetime*: 01/29-06/19 Wednesday,Sunday 6:00pm-7:00pm *Location*: Online10 Thanks, Sunshine Elite Education --------------------------- --------------------- Web: www.SunshineEliteEducation.com

1 0

exhibit: teacher to students, pre-class email, Silicon Forest May, 2022
by kirby urner 08 May '22

08 May '22

From: Sunshine Elite Education Date: Sun, May 8, 2022 at 12:07 PM Subject: Data Structure and Algorithms in Python: Sunday May 8: Pre Class Prim's Algorithm Greetings guys -- In the old days of purely paper textbooks, adding to them or changing them was hard and only happened from one edition to the next, with maybe years in between. Today, those of us using Jupyter Notebooks within a version control system for textbooks, likely have the ability to update them often. We now think in terms of an electronic textbook that has the ability to change continuously through frequent new "versions" (instread of through relatively infrequent "editions"). Along those lines... At https://github.com/4dsolutions/elite_school I have expanded ADS_sandbox_6.ipynb with new material regarding Prim's algorithm for finding the least costly spanning tree within a graph. Here's a Youtube I cite and include screenshots from: https://youtu.be/Uj47dxYPow8 Prim's Algorithm: Minimum Spanning Tree (MST) 371,043 views Jun 23, 2016 EducateYourself <https://www.youtube.com/channel/UC10JldFF3n7mRSGX6QfrQAA> 2.07K subscribers Short example of Prim's Algorithm, graph is from "Cormen" book. Another resource I recently added, a new link from our textbook: https://www.geeksforgeeks.org/5-best-books-for-competitive-programming/ Might these be books from your library if you see CP in your future? Are any of them online? Maybe you already have some data structure and algorithms books you would like to recommend and/or share about. I've included a few recommendations of my own e.g. Math Adventures with Python for example. I show a picture of its author, in front of my house. See you at 6 pm. Kirby *Course Name*: Data Structure and Algorithms in Python *Datetime*: 01/29-06/19 Wednesday,Sunday 6:00pm-7:00pm *Location*: Online10 Thanks, Sunshine Elite Education ------------------------------------------------

1 0

snapshot 2022: data structures and algorithms
by kirby urner 28 Apr '22

28 Apr '22

In a few minutes I'll join Sunshine Elite Education online and help out like a gym coach in some ways, because Algorithms and Data Structures, at the 8th grade level, is a doorway into a competitive sport: competitive programming (CP). Although the contests are geared towards youth, say high school level, in full adulthood comes the "coding interview" and so the same skills apply. Then later tonight, I'll be marching a small army of the the future through the numpy + pandas + matplotlib ecosystem. I'd call that Algorithms and Data Structures too, in the numpy's ndarray and the DataFrame are both data structures. With lots of algorithms to go with. My new Hello, World page: https://nbviewer.org/github/4dsolutions/clarusway_data_analysis/blob/main/K… My way of merging both worlds is to tackle the "longest increasing sequence in a matrix" problem, a coding interview classic, with numpy. I'm sure that would not be acceptable in the contest area, but it's a way of merging my two worlds. If you want to see more, click on USAOS in the Notebook allow and scroll to where I'm developing those algorithms (Kadane's Algorithm etc.): Eureka Moments. https://nbviewer.org/github/4dsolutions/elite_school/blob/master/Exercises.… I think of this as anthropology, me some guy in the field. Kirby

1 0

Fwd: exhibit: after class note to students, Silicon Forest elite school
by kirby urner 21 Apr '22

21 Apr '22

---------- Forwarded message --------- From: kirby urner <kirby.urner(a)gmail.com> Date: Wednesday, April 20, 2022 at 7:50:48 PM UTC-7 Subject: Re: exhibit: after class note to students, Silicon Forest elite school To: Trimtab Book Club <trimtab-book-club(a)googlegroups.com> A very similar letter to my recent one, why? I'm teaching two groups, this one entirely online. So what I test out with one group, I adjust for the next, leap frog fashion. I think it's worth sharing them both, to promote a stereoscopic view of the terrain. ======================== Thanks for tuning in to our biweekly "radio show" (like talk radio, but with pictures). Today our theme was algorithm speed, which is described using O-notation. "Polynomial time" O(n^2) for example, is only able to process about 5000 data elements in a second, rougly speaking. That's an approximation I cited from the USACO PDF I want you to have: https://usaco.guide/CPH.pdf That 300 page PDF has a healthy dose of what we call Discrete Math. It's what you will need to do well in CP (competitive programming). Sorting, finding shortest paths through a graph, or solving a subarray problem ("Kedanes algorithm" is part of the lore), have all been implemented for you. The USACO documentation counsels against trying to reinvent the wheel on the fly i.e. reinventing the best known algorithms. Rather, your job is to know which algorithm is called for, and then to make use of the relevant library, e.g. for sorting. My talk zoomed back, away from CP, to show more of the terrain: algorithms as used in the movie and simulations industries, data structures as tiles from a globe. I demonstrated a real time rendering of tensegrity structures, implemented by a world leader in the field, Gerald de Jong. Thanks to WebAssembly generated from Rust, your browser is able to show his creations involving in real time, not as a pre-recorded movie. Sometimes it's hard to tell the difference. https://pretenst.com/app/#construction;Triped https://pretenst.com/app/#construction;Halo-by-Crane Real materials (not CGI): https://pretenst.com/construct ion/2021/07/27/brass-bubble-120 <https://pretenst.com/construction/2021/07/27/brass-bubble-120> This is all still Algorithms and Data Structures. I ended with kennethsnelson.net. That's not the website I built for him, and for which he gave me Barrel Tower (showed on screen), but the one that came after, when he found the right paid professional. I was just doing it as a friend. Kenneth, an Oregonian (born and raised in Pendleton), was in inspiration for both Gerald and myself. Lastly, what some of you may of missed as I shared this early (as a loose end from last time): https://youtu.be/gFEE3w7F0ww Lex Fridman is a great interviewer and in this case his guest is a member of the Python community who has a lot to do with the tools I've been introducing (e.g. numpy and Anaconda). Kirby *Course Name*: Data Structure and Alg orithms in Python *Datetime*: 01/29-06/19 Wednesday,Sunday 6:00pm-7:00pm *Location*: Online10 Thanks, Sunshine Elite Education

2 1

update re Algorithms and Data Structures (pre college)
by kirby urner 21 Apr '22

21 Apr '22

What I didn't fully realize, when I signed on as the Algorithms and Data Structures teacher, for two groups of 8th graders, is that ADS is pretty much the property of CPers at the age level. CPers? Competitive Programmers. They join olympiads <https://nbviewer.org/github/4dsolutions/elite_school/blob/master/Exercises.…> , which present them with problems needing those special algorithms. There's a premium placed on speed, both to come up with a solution, to implement it, and to have it run at optimal speed. As the old guy applications programmer with a philosophy background (Princeton), I'm not a past champion. Nor have I had to go through a gauntlet of job interviews for coding jobs with big companies, and that's where CPers look next. Sure, maybe these olympiads are mostly for a younger crowd, but the ability to choose the right algorithm quickly is what will get them a job at the end of the day. That's the PR, which I'm just passing along. My initial approach was somewhat overlapping but different. I was more thinking in terms of *Mathematics* *for the Digital Age* and other discrete math tomes we have talked about at length in this archives. Working backwards from RSA (the public key algorithm), I know we want to cover Fermat's Little Theorem and Euler's generalization featuring totients and totatives. So what are those? Let's do some Group Theory while we're at it and show how the totatives of N, multiplied modulo N, form a finite group. Fibonaccis, generating primes, things of that nature <https://nbviewer.org/github/4dsolutions/elite_school/blob/master/ADS_intro_…> . Then, in a more innovative mode, I've worked to expand data structures to include "structured data" such as json, html, xml, csv i.e. serialization protocols. In addition, I've sought to bring in DataFrames from the array based computing world of data analysis. Numpy and pandas. I think for many in a high school setting, what will serve them going forward is good old Anaconda and all that implies. As a base of operations. As a main productivity tool. Jupyter Notebooks rock, in my opinion. What I'm attempting, in microcosm, is to make Algorithms and Data Structures useful to non-CPers as well, i.e. those not especially interested in or drawn to competitive programming. It's not like algorithms and data structures should go off their radar as a consequence. The last innovation I'm trying involves keeping things more right brain i.e. graphical not just lexical. A graph, or network, is already a visually accessible data structure. It's almost a no-brainer to then bring polyhedra into the picture, as more graphs, with edges and nodes. Polyhedra also need to be stored and retrieved somehow, i.e. back to structured data and SQL. Indeed, in other writings I advocate introducing polyhedrons as paradigm objects, when learning OO, as they're both abstract mathematical and colorful tangible. They have attributes. They come with methods (scale, rotate, translate...). Where do we go from polyhedra? To divided (tiled) spheres and global data. To world maps <https://nbviewer.org/github/4dsolutions/elite_school/blob/master/ADS_sandbo…>. More data structures. More algorithms. I started a new sandbox intending to import global data tiles, but that ended up corrupting my old Mac Pro. Anaconda snuck in some incompatible upgrades. So far, I only have one "sandbox" <https://nbviewer.org/github/4dsolutions/elite_school/blob/master/ADS_sandbo…> devoted purely to polyhedrons in the ADS portion of my repo (much more elsewhere), a table of relative volumes built with pandas. The volumes won't seem familiar and the relevance of this table is not widely echoed in other literature, except in the matrix of suppressed American philosophy. One of our leading futurists, from the 1960s through the early 1980s, made a big deal out of the tetrahedron and its ability to serve as a basis for our relative volume concept. Even a tetrahedron of edges one could have volume one, in his recasting. This is a topic I've posted about extensively here on edu-sig over the years, and on Math Forum, although the latter was purged of its public discussion groups, breaking lots of links. My attitude is I want to be one of those teachers who took the Bucky Fuller stuff seriously enough to want to include it in downstream published open source works. You'll see if you go to bfi.org/synergetics that my Github writings <https://nbviewer.org/github/4dsolutions/DigitalMathematics/blob/master/Mart…> are currently on tap to give scholars some exposure to the philosophy in question. Part of walking my talk is to showcase said material in a workaday context, namely in an 8th grade level Algorithms and Data Structures class. Kirby

1 0

a curriculum dev in the Silicon Forest
by kirby urner 03 Apr '22

03 Apr '22

I'm continuing to develop curriculum for a tiny elite college prep school smack in the center of the Silicon Forest, a place called Bethany Village, which specializes in offering after school elective time activities (e.g. martial arts, foreign language learning). [0] Nearby: Intel, Tektronix, ESI... [1] The Sunshine Elite School runs both after school enrichment and for-credit day school programs. I call myself "a developer" or even a "dev", but if people ask what I develop, I say "curriculum". I teach four times a week and look for ways to include student work in ways they permit. The school already has a Hall of Fame for competitive tournament winners. Doing a stellar project around Wordle or Chess would be another way to develop one's portfolio, through lasting links in the text itself. I wonder what other educators here think of the following proposition: === Most of the wood pulp published Algorithms and Data Structures (ADS) textbooks inherit from a day before "array based computing" and therefore have no "DataFrame" to integrate. Not as such. Of course they have 2D arrays (Python: lists of lists), i.e. row and column data, but not anything as sophisticated as a DataFrame in R or pandas. So let's include the DataFrame in a next iteration of ADS! That's certainly what I'm doing.[2] That puts machine learning algorithms within the reach of a pre-college student. We needn't dive deep into deep learning. We're free to dabble. We're still young and carefree (student attitude). "Buckle down data science" comes later, for those who so choose. === I see Jupyter Notebooks (Colab, Kaggle Notebooks etc.) as an extension of the desktop publishing revolution of the 1980s (PageMaker + Apple LaserWritier == professional quality publications). I remember working with nonprofits and NGOs and those days (at a place called CUE), and how excited these organizations were to be able to upgrade the look and feel of their literature. The web itself held a similar promise: the little guys wouldn't be trounced by the big guys, just because the big guys were big. Jupyter Notebooks, with embedded videos and data visualizations etc., put more of a level playing field under all would-be publishers. Our content is potentially as high quality as the content in professional papers and magazines. As long as we study our Tufte... Look how Jake VanderPlas goes back and forth between Notebooks and a wood pulp (non-interactive, yet standard setting) O'Reilly book. [3] As a curriculum dev, I'm able to bypass the wood pulp publishers entirely, using open source tools to create open source assets available to fellow faculty. To me, this looks like a quiet revolution against the hegemony of "those with enough money to dress up" (i.e. make their stuff look pro). Kirby Cc: Wwwanders (Silicon Forest, Doug Strain affiliate) Citations: [0] https://flic.kr/s/aHBqjzJdXN [1] https://www.oregonencyclopedia.org/articles/silicon_forest/ (mentions Doug Strain) [2] https://nbviewer.org/github/4dsolutions/elite_school/blob/master/ADS_sandbo… [3] https://github.com/jakevdp/PythonDataScienceHandbook

1 0

EliteSchool Repo meets Stanford GraphBase
by kirby urner 12 Mar '22

12 Mar '22

I was following the Wordle craze, looking for a list of five-letter words. Little did I know I'd stumble upon the Stanford GraphBase. https://github.com/4dsolutions/elite_school/blob/master/ADS_research_1.ipynb There's also some Pi Day stuff, for teachers taking advantage of 3-14. Kirby ==== Exhibit from March 2022: Hi Guys -- Our topics today included: * arbitrary precision numbers as provided by decimal.Decimal or gmpy2 * algorithms for converging to Pi (one of Ramanujan's and one I learned from Guido) * another rear view mirror look at Five Letter Land * SymPy https://mpmath.org/ Friendly Repo: https://nbviewer.org/github/4dsolutions/Python5/blob/master/Pi%20Day%20Fun.… Friendly Replit: https://replit.com/@kurner/piday#main.py In fact we were indeed discovering some of the same Graph Theory facts about the list of five letter words, that come to light in the Stanford GraphBase course, authored by Knuth himself (Chinese name "Gartner"): https://blog.krybot.com/a?ID=0 1800-af2c69d8-40ba-4b02-9cdc-aca2c205be68 <https://blog.krybot.com/a?ID=01800-af2c69d8-40ba-4b02-9cdc-aca2c205be68> https://replit.com/@kurner/fiveland#main.py I finally got that Regular Expressions sandbox to work: https://regex101.com/r/k0eF6r/1 I felt it important to bring up the topic of Regular Expressions because they're used in many algorithms where the task is to do pattern matching. See you Monday. Keep thinking about those projects. Mandelbrots and Mandelbulbs: another topic we touched upon. Kirby *Course Name*: Data Structure and Algorithms in Python *Datetime*: 01/29-06/19 Wednesday,Sunday 6:00pm-7:00pm *Location*: Online10 Thanks, Sunshine Elite Education ------------------------------------------------

1 0

pyalgoviz solely in-browser solution?
by Jurgis Pralgauskis 19 Feb '22

19 Feb '22

Hi, I stumbled upon old good https://pyalgoviz.appspot.com/ I wonder what are options to do it solely in browser: - skulpt? - brython? do they support `sys.settrace` functionality -- maybe with some js under-the-hood? -- Jurgis Pralgauskis tel: 8-616 77613; Don't worry, be happy and make things better ;) https://galvosukykla.wordpress.com/

1 0