Print Story stuffing things. also: ChiHuSiMeet!
By gzt (Tue Apr 17, 2012 at 01:31:41 PM EST) gzt, chihusimeet, poutine, data, analytics (all tags)
I had a slight "D'oh!" moment yesterday when I was glancing the website for a conference on topological data analysis and noticed there were two people from OSU who attended. I hadn't noticed them earlier because they had courtesy appointments in stats, so I didn't really look at their work. Well, I still prefer my choice and I don't think that's the direction I will go, but that's the sort of thing that would have moved them a notch higher than they were in terms of "research fit". Still, I don't like making decisions and then finding something right afterwards that I deem relevant.

INSIDE: more Windy City Husi Meetup discussion.

Bad Happy Poutine tonight. They have a challenge thing. I'm totally doing it. Anybody else is welcome to google it and drop by at around 6. If you really want, you can drop by church at 5, too.

Watched more of The Wire last night. It's still awesome.

I think tonight I'll finish my stats project and do my stats homework. Well, might not finish the project, but will get a start drafting. Then I'll spend the rest of the week doing numerical analysis. Projects for both due 4/25. I'm kind of interested in going to this: but I really need that day to work on projects. But, honestly, I wouldn't quite fit in, not being an actual user of Big Data at the moment and not knowing much at all about Hadoop. Still, it's hard to beat "free", networking is good, and the subject is, ostensibly, one I'm going to be interested in.

ChiHusiMeet: looks like we're going to do this in some form. Unless we get a strong call for a multiday event, here is my suggestion: find somebody's house/condo/apartment/park and grill or otherwise consume the finest meats and cheeses of the land on a Saturday afternoon. For multiday people, we could surely find something interesting to do on Friday evening and Sunday afternoon, possibly even Saturday morning as well (watching footie at The Globe, perhaps, because this is nominally a UKian site). At present, it seems like on the order of 10ish people and is therefore feasible in many locations.

I did not sense a strong consensus or preference on dates, so we need to continue the discussion. I'm still not sure what dates work yet, either, so this isn't yet a priority. I think we should lean toward June or July rather than May?

< not a lot to say | The World's Biggest Biplane >
stuffing things. also: ChiHuSiMeet! | 10 comments (10 topical, 0 hidden)
I would by barooo (2.00 / 0) #1 Tue Apr 17, 2012 at 02:27:52 PM EST
vote for a trek out to the 3floyds pub because it's world class.  But it's also a clusterfuck.  A group of indeterminate size on a weekend would have a wait of approximately 17 hours.

I would also offer to host but our place is tiny and whilst I have grilling and barbecuing experience, we just don't have that kind of size.  Nor would the wife be keen on hosting a bunch of people she doesn't know, causing undue stress.

But I would be willing to assist in grilling activiities and probably even fund some meat purchases. 

man, i need a beefy taco now.
My place is nice and can party well with about 12 by gzt (2.00 / 0) #2 Tue Apr 17, 2012 at 02:41:51 PM EST
If people are loitering long term. Short-term partying, of course, can have quite a few more, since you aren't having to worry about seating so much.

There a few other pretty good breweries around, perhaps we could research whether any of them are not such a clusterfuck?

[ Parent ]
by that i mean... by gzt (2.00 / 0) #3 Tue Apr 17, 2012 at 02:50:45 PM EST
...i have, without shoving anything around and digging out chairs, natural seating for 12. If people are standing up and milling around, I think 16 is doable for a few hours and 20 for a short period if I rearrange things - more if people go on the deck. I've got a pretty open floorplan. But I have to contend with my wife and a roommate - I think they would be cool, though.

[ Parent ]
Revolution is also probably a mess by LoppEar (2.00 / 0) #4 Tue Apr 17, 2012 at 02:52:59 PM EST
Although if we're less than 10 it might be worth calling if that's what we want to do. I'd rather not do Goose Island.

Hopleaf, though not a brewery, does have a very excellent beer selection (heavy on belgians) and is currently doubling their space, looks like they'll be finished in the next few weeks (and conveniently just around the corner from me ;) )

I also like the Globe suggestion.

If we're available to host, sounds like our apartment is about the same size as yours - we've done parties with ~15.

[ Parent ]
The Hop Leaf is pretty great by gzt (2.00 / 0) #5 Tue Apr 17, 2012 at 03:06:08 PM EST
The Globe is pretty good for drinking on a Saturday morning, but Hop Leaf shines in general.

Do you know anything about Half Acre? I also heard there's a place up in Ravenswood that also started distilling or something like that...

[ Parent ]
Koval is the distillery by LoppEar (2.00 / 0) #6 Tue Apr 17, 2012 at 03:24:03 PM EST
It's a tiny place, worth a tour from the friends who've gone but I'm not sure they can even accommodate a tour of 10.

Half Acre, I haven't been, it's just a brewery tour I thought (no pub?). Haymarket is on par with Revolution.

[ Parent ]
And if we're by barooo (2.00 / 0) #7 Tue Apr 17, 2012 at 03:27:39 PM EST
not limiting ourselves to actual brewpubs, I like the Fountainhead.  Plus, it's easy staggering distance from my place.

Revolution is good too.  As is the globe, which I can also walk to but not quite as easily (3/4 mile vs like 2 blocks).

Half Acre makes good stuff and I think they do tours but you can't drink there.  City Provisions (by me) had a tasting of the distillery's stuff at one point I think. 

man, i need a beefy taco now.
[ Parent ]
i've said it before, by the mariner (2.00 / 0) #8 Tue Apr 17, 2012 at 03:44:35 PM EST
but i'll say it again: i'm pretty sure most of this topological data analysis stuff is bullshit and it will turn out to be a fad that amounts to a way to separate the nsf and various firms with mediocre analysts (or managers who read more than they should) from their money, but we'll see. 

could very well be by gzt (2.00 / 0) #9 Tue Apr 17, 2012 at 04:02:05 PM EST
but it's mildly intriguing, enough to glance at as a possibility if I were some place that did it.

firms with mediocre analysts, though, are at least smart enough to notice if their hoity-toity topological analytics are earning them money. and they're typically following the trend of more successful businesses rather than investing in tools like that. no, i think it's high-end firms that will get suckered in.

[ Parent ]
can't knock until you've tried it by the mariner (2.00 / 0) #10 Tue Apr 17, 2012 at 09:49:32 PM EST
or until the manager who suggested using it gets promoted past caring about it. probably true that high-end firms will use it, others will either try it or wish they could afford to try it.

i agree it sounds cool. i should qualify: what i think is probably going to turn out to be useless bullshit is this persistent homology stuff and related techniques. "topological data analysis" sounds broader and i haven't looked into it enough to know how much broader it is. wikipedia seems to suggest not that much. it just seems to me that in real life, quantitative things are what matter, not things like whether there are cycles that stay near your data set that don't bound some other chain that also stays close to the data set. (okay, so a betti number is a quantity, but it describes how many cycles like the above are around that are linearly independent up to equivalence by bounding some chain that stays close to the data, blah blah blah. maybe the filtration they put in there gives enough control over the geometry to extract nice quantitative stuff, but i haven't seen/heard too much evidence of it.) abstracts for talks about this stuff i see usually look full of buzzwords.

[ Parent ]
stuffing things. also: ChiHuSiMeet! | 10 comments (10 topical, 0 hidden)