suggested computer language for explaining some algorithms

[bhelmes]

A peaceful and pleasent night for you,


what kind of computer language would you suggest for a mathematical book about number theory in order to illustrate the algorithms ?


The programming language should be

a) easy to understand,
b) supported for a long time (20 years at least) and
c) usable for different operation systems


I used Mupad some years ago, but it is vanished.


Recommandations ?


Greetings from the polynomial rings

Bernhard

[chalsall]

Originally Posted by bhelmes

Recommandations ?

While I personally can't stand it (it uses indentation for code blocks), Python would probably be a good choice.

[VBCurtis]

Perhaps psuedocode for illustration, but actual code isn't called for in a math text.

If you can't explain the algo in human language, your exposition needs work.

[hansl]

PARI/GP is convenient for math and number theory related algorithms.

[kriesel]

Just bouncing some ideas around here:
natural language, constructs like if then, else, until, while, for, and parentheses;
whichever language more readers of the book are likely to already know;
perhaps vary the language among a very small set according to which seems to best suit the specific situation;
something included at Rosetta Code;
something available at no additional cost or very low for your students if this is for teaching;
Wolfram (Mathematica) but very costly;
stay with what you know; MATLAB has an optional toolbox with MuPAD (and interfaces with various other languages) https://en.wikipedia.org/wiki/MATLAB.
MATLAB student cost doesn't look too bad.

[LaurV]

Pascal.

No (big) difference from pseudocode, human-like, easy to explain any algorithm in it, easy to be re-written in any other language. In C, you can easy compile most of Pascal source code just #defining "begin" to "{", end to ";}", and a couple of other tricks.

Technically, if you have the logic diagram (with blocks) in front of you, or in your mind, writing Pascal programs from it is just a piece of pizza...

After learning Fortran and Basic in high school (for Fortran we used punched cards on a Felix C256 our highschool owned, and for Basic it was the early days of the intel 8080, etc and we had a M18 with centronix console - i.e. printing on paper, this was earlier than the displays and terminals era), we went to uni where they used to teach us Pascal (on Coral, - you MUST click this link! - and serial DAF2020/VT100 terminals in all the building, no listing papers and punch cards anymore), and it was such a revelation... (we studied C started from the second year at uni).

[kriesel]

Originally Posted by LaurV

Pascal.
...
After learning Fortran and Basic in high school (for Fortran we used punched cards on a Felix C256 our highschool owned, and for Basic it was the early days of the intel 8080, etc and we had a M18 with centronix console - i.e. printing on paper, this was earlier than the displays and terminals era), we went to uni where they used to teach us Pascal (on Coral, - you MUST click this link! - and serial DAF2020/VT100 terminals in all the building, no listing papers and punch cards anymore), and it was such a revelation... (we studied C started from the second year at uni).

Pascal occurs to me as a language with a great past but not so much a future.
I first learned Fortran in the wait-for-a-keypunch-to-be-free days, then BASIC, various assemblers, c, Pascal, others. Pascal seemed a step up from Fortran toward natural language, and c a step in the other direction toward assembler. https://en.wikipedia.org/wiki/Pascal...mming_language)

[Happy5214]

Current college students have virtually no experience with Pascal (which is a shame, since I've heard great things about it). One of my professors was a huge fan of PL/I, lamenting its loss to C.

I think you should probably use pseudocode when targeting a non-CS audience. If it needs to be an actual, modern language, Python and Ruby are readable, cross-platform languages you can use.

[xilman]

Originally Posted by bhelmes

what kind of computer language would you suggest for a mathematical book about number theory in order to illustrate the algorithms ?

If you need any specific programming language to explain your algorithms, you have failed to explain your algorithm properly. There are far too many language-specific irritations in any given language.

Almost any form of pseudo-code should be sufficient. That used by Knuth in TAOCP is a classic of clear and precise specification.

[kriesel]

Originally Posted by xilman

Almost any form of pseudo-code should be sufficient. That used by Knuth in TAOCP is a classic of clear and precise specification.

MIX?!


Nice avatar. But be careful of using it in real life. Around here, 4 could get you 0 to the chin.

[bhelmes]

Originally Posted by VBCurtis

Perhaps psuedocode for illustration, but actual code isn't called for in a math text.

If you can't explain the algo in human language, your exposition needs work.

Perhaps my english knowledge is not perfect, and may be the chinese reader will not detect every thought, therefore a second explication could improve some knowledge transfair easier.