sje30 / catam-julia Goto Github PK

@jmb280cam in the section "Learning Julia and its advantages"

with the bullet points to references, could you add the following two points somewhere

1. The Computational Thinking course at MIT provides an excellent set of videos to introduce julia to those new to programming. https://computationalthinking.mit.edu/

2. the Julia Express https://github.com/bkamins/The-Julia-Express/ is a useful reference guide to the core features of Julia.

hi @jmb280cam

here is a list of ideas for improvements from Gokul @srgk26

1 how automatic differentiation works (not just how it's implemented, as in the video by Alan Edelman)
2 types and type stability; also highlight difference between type annotation and type stability (the former does not improve
 performance)
3 refer to blog for type stability in julia: https://www.juliabloggers.com/writing-type-stable-julia-code/
4 using @code_warntype on function calls to check for type stability
5 custom typing with struct
6 multiple dispatch
7 highlight performance tips page in docs: https://docs.julialang.org/en/v1/manual/performance-tips/
8 using maths macros @simd, @fastmath, @inbounds to further improve performance (not always, can reduce performance -
 case-by-case basis)
9 good idea to introduce docker as well
10 tutorial for common tools and packages in julia ecosystem (dataframes, plotting, flux, differentialequations, etc.)
11 note that julia is column-major, so iterate through rows before columns in matrices, for example:

m = rand(2,3)
@inbounds @views for j in 1:size(m,2)
    for i in 1:size(m,1)
        m[i,j] = m[i,j]*(i+j)
    end
end

some of these could be in the intro, some (like types) would be better off in a case study.

CC: @Nick-Gale for info

I've changed the format of the readme.md for the Permutations case study, see here. I thought it would be better to have all of the links relating to the project in one place so that students could get all of the files without having to go through GitHub.

Any thoughts on this or other changes that should be made @sje30 @valentin-imbach? I'll look to change the rest of them later to match after your feedback.

How about using the preferred method for documenting functions?

https://docs.julialang.org/en/v1/manual/documentation/

e.g. "..." or """....multieline""" before the function?

hi @jmb280cam I think there is a typo in the intro doc; seperate should be separate

hi @jmb280cam

the section https://sje30.github.io/catam-julia/casestudies/voronoi/voronoidiagrams.html#5619eeaa-d0ae-4752-987a-abc6b9236725 would be easier to understand I think with a figure to help demonstrate how your algortihm works? Could you sketch one? (either just in a png or using Plots() to draw it.)

fonsp/Pluto.jl#1309 (comment) is a temporary fix
(adding the html block) to work around the scrollbars on the right of paragraphs that have mathml on the last line of the paragraph.

Please can one of you check to see if that works? @valentin-imbach @jmb280cam

I've started working on a Voronoi diagram case study, but I'm not sure how I could turn it from a computationally heavy task into one with a mathematical purpose. Does anyone know of an application / problem that can be solved which I could incorporate?

fonsp/Pluto.jl#1149 (comment)

Some of these may be fun to convert to Julia,

https://graphics.stanford.edu/~seander/bithacks.html

fonsp/Pluto.jl#1314 shows that we can edit the .jl file.

compiling it on the go using: Pluto

fonsp/Pluto.jl#1149

there is already a python package to do what we discussed @valentin-imbach @jmb280cam

https://github.com/LeviBorodenko/primify

this prime drawing looks nice, but perhaps we can use Julia to do it a bit smarter /more efficient than what can be done in Python?

https://github.com/LeviBorodenko/primify
https://esoteric.codes/blog/primify

or, synthesising an image from some text (like a CRSid) and then finding the prime? as I said, it might be fun for outreach days, when young kids can come along and get their name converted into a prime number!

hi @jmb280cam the SIR model looks great. just one minor suggestion right now, red/green/blue is not a good combination -- 7% of males are r/g colorblind. can you try another triplet of colors? If short of ideas, colorbrewer.org is good, e.g. https://colorbrewer2.org/#type=qualitative&scheme=Dark2&n=3

Hamza has suggested prioritising the following three folders:

"interp", "ode" and "lu_fac"
from

https://www.damtp.cam.ac.uk/user/naweb/ib/

access to matlab might be needed to check what the GUI in these folders is doing, and how we might best circumvent a GUI.

Here are some options for formats that case studies can be written in;

Pluto

• Positives
  • Good interface, outputs clean webpage or interactive notebook
• Negatives
  • PDF output is not great
  • Not usable for all applications, since cannot have WIP code or reassign variables

Jupyter

• Positives
  • Good output and interactivity like Pluto, and can reassign variables
• Negatives
  • A pain to use (at least as I've found so far)
  • Still can't have WIP code

LaTeX

• Positives
  • Looks mathsy
  • PDF output
  • Images/PDF outputs can be naturally imported
• Negatives
  • No interactivity
  • Slower to write than Word/Google Docs

Word

• Positives
  • Can have a PDF output
  • Can copy code in from VSCode and keep the colours
• Negatives
  • No interactivity
  • Lacks the precision of LaTeX
  • No live collaboration like Google Docs
  • Doesn't look like LaTeX, might be less appealing

Google Docs

• Positives
  • Can have PDF output
  • Can copy code in from VSCode and keep the colours
  • Multiple people can work on it at the same time (may not be necessary)
• Negatives
  • No interactivity
  • Lacks the precision of LaTeX
  • Less features than Word
  • Doesn't look like LaTeX, might be less appealing

Whether we choose between them on a case (study) by case (study) basis or pick one or two to use for everything I don't know. We should probably also create quick templates/style guides for those that we will use for consistency

Having pluto code run in mybinder is great, but how is support for pluto in cocalc?

this thread seems relevant:

sagemathinc/cocalc#5270

We can get a licence for cocalc if that is helpful to explore.

perhaps we need to have a section on 'introspection' to show how to examiner objects and methods?

e.g. for starters, @edit(sin(1.0)) to show the code for how the sin of a double precision number is stored

and something like this for understanding different types of Integers

x = 64
typeof(x)
bitstring(x)
y = Int8(x)
bitstring(y)

with floats we could do it after describing the representation https://en.wikipedia.org/wiki/Single-precision_floating-point_format but that's also feasible.

random walks looks great @jmb280cam

I tried to run it via mybinder though and got error

ArgumentError: Package Plots not found in current path:

it has presumably not loaded any of the packages in mybinder to run? is it because the files are split?

hi @jmb280cam
please can you upload the introduction.jl (and perhaps the html output from pluto) into the intro/ folder here on github?