Want to take part in these discussions? Sign in if you have an account, or apply for one below
Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.
1 to 19 of 19
I dropped some query boxes on some basic questions and observations over at locally presentable category. I ask experts to help bring me up to speed.
Not that I've answered your questions, but I consolidated the discussion on terminology there. (Mike and Urs may want to note that Reid Barton responded in December to Mike's final point, so the discussion is still open.)
Thanks, Todd, for these remarks. Yes, given what i understand I would think that the answers are Yes, notably to the first query box. Notably that would then harmonize with the statement for oo,1-categories.
But I would like to cross-check this with exactly how this is stated in "Locally Presentable and Accessible Categories". I had a Google books copy of that book with some of the pages available, and what I did put into this entry I gotout of these pages that I saw on Google books. It was a bit of a pain looking at the the book that way, and I can't claim to have a good overview of the content of that book. Given how straightforwward the statements you suggest should be true are, one would think they'd be mentioned this way in the book, but I can't remember having seen them. So I would just like to cross check again. Possibly they have some extra funny condition of what counts as a sketch for them?
Trouble is, now that I search around, I can't even seem to find the Google-books version of the book. (??)
Thanks, Urs. Yes, I can't find the Google Book either. But I think I have some answers. There is in particular an important theorem found in the book on accessible categories by Makkai and Paré (which I don't have), to the effect that the 2-category of accessible categories and accessible functors admits weighted bilimits. This implies in particular that if $A$ is locally presentable and $C$ is small, then $A^C$ is locally presentable, which answers one of my questions. I'm not yet sure what the answer is to this in the locally finitely presentable case.
Regarding the question about reflective subcategories, the impression I'm getting is that the answer is 'no' in general; the answer would be 'yes' if the inclusion is an accessible functor (which I think would boil down to the inclusion preserving $\kappa$-filtered colimits, if the containing category is $\kappa$-accessible). This 'yes' is actually another corollary of the Makkai-Paré weighted bilimit result; more generally, the Eilenberg-Moore category for an accessible monad (on an accessible category) is accessible by this result.
I found all this in an online copy of Tibor Beke's thesis. I would still like to be further educated by experts around here.
Hi Todd,
good. Would you have the time to put in the answers to your questions, as far as you have them now, into the entry? Or should I do it?
By the way, to get math displayed here, it needs to be enclosed in a pair of double dollar signs. (And may not have any line breaks in between these.)
No, I can do it Urs. Thanks for the offer.
I knew about the double dollars here, but forgot because I was in a hurry. Thanks though.
Okay, removed the old query boxes, added some content, and also added some new query boxes (which are mostly notes to self) to locally presentable category.
Replied to Reid.
Todd,
thanks, this is a major improvement to what we had. Nice. It's very good to have you around.
Asking a question here, because it is slightly quicker.
Can the 'locally' in locally presentable category refer to properties of the slice categories C/a or a/C? Or more precisely, can a locally presentable category be defined as a category C such that the slice categories C/a (or a/C, as necessary) have such and such a property?
@David: I don't think so.
I made a small fix - the κ in "locally κ-presentable category" doesn't mean that all objects are κ-compact, only that the generating set can be taken to be so.
Ok thanks
for completeness, I have added to locally presentable category what should have been there but wasn’t: the remark after the statement about sheaf toposes being presentable that it’s of course the set of representables that serve as generators, being -small for bounded below by the cardinality of morphisms of the site (now prop. 6 there).
@Todd (#4):
I’ve added a proof that is locally finitely presentable when is small and is a locally finitely-presentable category.
Thanks. I have added some links.
Thanks, Zhen! (At the time of writing #4, I was evidently far less fluent in this part of category theory than I feel nowadays! (-: )
Edit: And now I just tried to look and see the proof you wrote, but I can’t find it anywhere. I didn’t see it under Recently Revised.
Edit: Ah, there it is – never mind. Thanks again.
I changed “finitely-accessible” to link to accessible functor instead of accessible category. Basically, I’m trying to say that the inclusion is finitary… but the wording is ambiguous and should be improved.
Thanks, right. I wanted to link to something like accessibly embedded subcategory, but we don’t have such an entry. We should, though.
1 to 19 of 19