Name for Category Theory Module

[Sergiu Ivanov]

Hello,

I am trying to figure out what name to give to the new category theory
module. I have considered a number of possibilities:

* ct -- this corresponds to the common abbreviation of "category
theory" as CT, but I'm afraid is absolutely non-suggestive;

* categories -- this seems to fit nicely with the nomenclature of
existing modules in SymPy (polys, series, integrals, matrices,
etc.), however, I am not sure as to the intuitiveness of this name
either; I wonder how easy it would be to confuse this name for
something having to do with categories in the sense of groups
resulting from a classification?

* cattheory -- looks the fairest of all, however, it looks somewhat
ugly to me (the double "t"); catheory looks totally abominable;
catstheory sounds better when read, but looks even uglier;

* categorytheory -- the fairest and the easiest to read, but it might
be a bit too long; it's my favourite though.

Which one would be more preferable in the opinion of the SymPy
community?

Sergiu

[krastano...@gmail.com]

I would like an underscore there: category_theory

And shamelessly hijacking the thread: are there any objections to
`differential_geometry` for the module on which I am working?

[Chris Smith]

> * categories -- this seems to fit nicely with the nomenclature of
>  existing modules in SymPy (polys, series, integrals, matrices,
>  etc.), however, I am not sure as to the intuitiveness of this name
>  either; I wonder how easy it would be to confuse this name for
>  something having to do with categories in the sense of groups
>  resulting from a classification?

Anyone interested will learn quickly what it stands for.

> * categorytheory -- the fairest and the easiest to read, but it might
>  be a bit too long; it's my favourite though.

This is ok, too. ntheory is used for 'numbertheory' so maybe ctheory
could be used?

[Tom Bachmann]

Hi,

I have no strong opinion and I don't think it matters much either way.
If we come up with an ingenious choice later (before the first release,
obviously), renaming is always a possiblity.

I agree ct is too short. Imho it shouldn't be too long either (because
then I have difficulty thinking the module name in my head -- not sure
if this explanation makes sense or is even relevant).

My personal favorite would be categories, because it is in keeping with
other sympy module names. This would suggest the unconventional
"manifolds" for the differential geometry module.

In any case, my suggestion (to both of you) is to not waste much thought
on the module name. It is used very seldomly in code anyway (because in
sympy we tend to "import the leaves").

Best,
Tom

[Sergiu Ivanov]

On Wed, May 23, 2012 at 6:35 PM, krastano...@gmail.com
<krastano...@gmail.com> wrote:
> On 23 May 2012 17:25, Sergiu Ivanov <unlimite...@gmail.com> wrote:
>>
>> * categorytheory -- the fairest and the easiest to read, but it might
>>  be a bit too long; it's my favourite though.
>
> I would like an underscore there: category_theory

I don't see many underscores in SymPy module names, besides, it makes
the name yet longer. However, it looks rather fine on the overall, so
if people vote for this, so be it :-)

> And shamelessly hijacking the thread: are there any objections to
> `differential_geometry` for the module on which I am working?

This name would make category_theory look perfectly in place ;-)

Sergiu

[Sergiu Ivanov]

On Wed, May 23, 2012 at 6:37 PM, Chris Smith <smi...@gmail.com> wrote:
>
>> * categories -- this seems to fit nicely with the nomenclature of
>>  existing modules in SymPy (polys, series, integrals, matrices,
>>  etc.), however, I am not sure as to the intuitiveness of this name
>>  either; I wonder how easy it would be to confuse this name for
>>  something having to do with categories in the sense of groups
>>  resulting from a classification?
>
> Anyone interested will learn quickly what it stands for.

Good point :-)

>> * categorytheory -- the fairest and the easiest to read, but it might
>>  be a bit too long; it's my favourite though.
>
> This is ok, too. ntheory is used for 'numbertheory' so maybe ctheory
> could be used?

I was thinking of this possibility, but I am totally unsure as to it,
so I'll wait for votes to collect.

Sergiu

[Sergiu Ivanov]

On Wed, May 23, 2012 at 6:45 PM, Tom Bachmann <e_m...@web.de> wrote:
>
> I have no strong opinion and I don't think it matters much either way. If we
> come up with an ingenious choice later (before the first release,
> obviously), renaming is always a possiblity.

Sure.

> My personal favorite would be categories, because it is in keeping with
> other sympy module names. This would suggest the unconventional "manifolds"
> for the differential geometry module.

That's the feedback I expected :-)

> In any case, my suggestion (to both of you) is to not waste much thought on
> the module name. It is used very seldomly in code anyway (because in sympy
> we tend to "import the leaves").

I'll wait for a little while to see if there are other suggestions; if
nothing new comes up, I'll stick with "categories" this time (since I
expect that more people would vote for it).

Sergiu

[Joachim Durchholz]

Am 23.05.2012 17:37, schrieb Chris Smith:
> This is ok, too. ntheory is used for 'numbertheory' so maybe ctheory
> could be used?

That naming scheme will give "gtheory" for both group theory and game
theory.
For that reason, I consider "ntheory" a naming mistake best not repeated.

[Joachim Durchholz]

Am 23.05.2012 17:45, schrieb Tom Bachmann:
> I agree ct is too short. Imho it shouldn't be too long either (because
> then I have difficulty thinking the module name in my head -- not sure
> if this explanation makes sense or is even relevant).

Yes, it does make sense and yes, it is relevant.
If something is inconvenient to call up to conscious thought, thinking
about it becomes cumbersome. (Not an original thought. I think Edsger
Disjkstra popularized the idea.)

[Aaron Meurer]

How about diffgeo, or diffgeometry?

Aaron Meurer

[krastano...@gmail.com]

>>
>> And shamelessly hijacking the thread: are there any objections to
>> `differential_geometry` for the module on which I am working?
>
> How about diffgeo, or diffgeometry?
>

diffgeometry sounds a bit better than the current one. I will probably
use it. I was also amused by the option to call it 'manifolds'.

[Vinzent Steinberg]

Am Mittwoch, 23. Mai 2012 21:52:32 UTC+2 schrieb Aaron Meurer:

How about diffgeo, or diffgeometry?

I like diffgeo most, because it's short.

Vinzent

[Chris Smith]

We use "geo" for geography (which I know doesn't apply to a CAS) but
perhaps it could be diffgeom?

[john.hoebing]

On May 23, 1:09 pm, "krastanov.ste...@gmail.com"

If we're going to shorten 'differential_geometry' (I'm also OK with
spelling it all out) then let's use a more canonical shortening. I
think the most common is 'diffgeom' which is more popular than
'diffgeo' or 'diffgeometry'; its more like 'diffeq' vs
'differential_equations'. You can google 'diffgeom' and see that its
used quite a bit in math departments when they need an abbrv.

John Hoebing

[krastano...@gmail.com]

I will probably end using that last suggestion: diffgeom