if you were writing a physics paper with a lot of QM,

[Mike Marchywka]

On Thu, Jan 05, 2023 at 02:23:39PM +0000, William F. Adams via texhax wrote:
>    Would it be an option to use literate programming?

https://en.wikipedia.org/wiki/Literate_programming

>    That allows one to use the full power of TeX (in making the PDF) and to write out arbitrary files (which would be your
>    computer algebra system), and to get a nice typeset .pdf
Thanks. I guess there are several issues to consider
for author/researchers. My concern is just avoiding mistakes
and that includes in algebra. Keeping things as "machine readable"
as possible is probably one goal then. There are a lot of things
you could do with equations but algebra, input to high performance
code, and publishing come to mind. And then of course for publishing
getting the output of the "high performance" stuff is maybe
another issue. 

Structured documents more generally may be an interesting 
thing to explore. I guess you could almost use latex
as a source document or notebook if its written
right. I could look more seriously at parsing latex equations
into something you could evaluate.  


>    William
>    -----Original Message-----
>    From: Deyan Ginev <deyan.ginev at gmail.com>
>    To: Mike Marchywka <marchywka at hotmail.com>
>    Cc: texhax at tug.org
>    Sent: Thu, Jan 5, 2023 9:15 am
>    Subject: Re: if you were writing a physics paper with a lot of QM,
>    Hi Mike,
>    If you wanted to integrate your numerical tools with your publication tools, the sane choice would be as you mentioned -
>    have your CAS-powered notebook emit a LaTeX document as an export.
>    Sadly, that often turns out to be too "vanilla" for power users, so you may end up writing templates/customizations to the
>    emitted LaTeX. I will go on a limb and state that the perfect system for the workflow you describe has not yet been
>    created, though various interesting experiments are ongoing.
>    As one example, Google's "latexify" annotations for Python have been getting a positive reception recently:
>    [https://github.com/google/latexify_py]https://github.com/google/latexify_py
>    As to the article you used as an example, the TeX it was written in seems to already be friendly enough to do more than
>    just PDF. Here's a sample:
>    [https://ar5iv.labs.arxiv.org/html/1810.11016]https://ar5iv.labs.arxiv.org/html/1810.11016
>    Greetings,
>    Deyan
>    On Thu, Jan 5, 2023 at 8:43 AM Mike Marchywka <[mailto:marchywka at hotmail.com]marchywka at hotmail.com> wrote:
> 
>      I just ran into this,
>      [https://arxiv.org/pdf/1810.11016.pdf]https://arxiv.org/pdf/1810.11016.pdf
>      which seems to illustrate a lot of typical math.
>      If you were authoring math oriented
>      documents and wanted to integrate your analysis
>      and publications, what would your source code
>      be for equations like this? I would imagine
>      you could do it in mathmatica or
>      matlab and export latex but how does
>      that integrate with high performance
>      code that you may use for numerical
>      solutions?
>      Again, I'm not big on appearance and don't
>      expect my latex code to be publication quality
>      but just curious about maintaining consistency
>      among pieces to avoid publication errors.
>      And trying to figure out what software to work
>      with :) My algebra has not gotten much better
>      with age...
>      "MikeMath" or my home brew mathmatica
>      seems to be coming together- although
>      an "on the fly" incremental math renderer
>      may be better than latexmk. I'm trying to
>      figure out what to do with some tensor
>      and operator implementations as well as the
>      bra/ket notations etc.
>      Also finding bugs in "chromate" by TooBib
>      and stand-alone file downloader that uses
>      largely headless chrome. I should replace the
>      nodejs wscat with a real websocket library
>      and find a better way to serialize asynchronous
>      events but as kluged as it is is seems to work
>      pretty well for now. Hopefully as I encounter
>      more websites I can learn more javascript
>      for stuff to emulate. As this does not use
>      puppeter, and talks directly to the chrome debug port,
>      its interesting webtool lol.
>      Thanks.
>      --
>      mike marchywka
>      306 charles cox
>      canton GA 30115
>      USA, Earth
>      [mailto:marchywka at hotmail.com]marchywka at hotmail.com
>      404-788-1216
>      ORCID: 0000-0001-9237-455X

-- 

mike marchywka
306 charles cox
canton GA 30115
USA, Earth 
marchywka at hotmail.com
404-788-1216
ORCID: 0000-0001-9237-455X