|
Ochowie posted:Is there a c# or java equivalent to numpy/scipy/pandas? I would prefer to use either of those languages but the python numerical analysis stack is too good. http://fsharp.org/guides/math-and-statistics/ might interest you quick googling also says that ironpython used to have numpy/scypy ports (which could be used by other clr languages), but those projects have been abandoned for years and don't seem to be usable anymore
|
# ? Dec 15, 2015 23:12 |
|
|
# ? May 26, 2024 06:02 |
|
F# or like BoA go for Iron Python,
|
# ? Dec 15, 2015 23:22 |
|
rrrrrrrrrrrt posted:fyi this isn't the case except maybe at limited forms of heads up poker. wonder if anyones ever tried to hook up a kinect or w/e to a poker bot to read facial info like hook this research up to a poker robot
|
# ? Dec 15, 2015 23:29 |
|
rrrrrrrrrrrt posted:fyi this isn't the case except maybe at limited forms of heads up poker. That's kind of neither here nor there compared to predicting messy stuff like sports victories. Poker is fundamentally a solvable turn-based game where you can assign an expected value to any given action and map everything in advance, unlike basketball.
|
# ? Dec 15, 2015 23:31 |
|
Bloody posted:math.net has some stuff. what are you looking for specifically? Basic data analysis, regression analysis. Also, some interpolation methods and finite difference methods. Bloody posted:alternatively if you're at work or something ILNumerics looks like it could be legit Yeah this is all for hobby level stuff so this probably wouldn't work.
|
# ? Dec 15, 2015 23:47 |
|
Ochowie posted:Basic data analysis, regression analysis. Also, some interpolation methods and finite difference methods. rapidminer?
|
# ? Dec 15, 2015 23:51 |
|
Jeffrey of YOSPOS posted:I want to play machine learning games which finance is ripe for - I'm not sure gambling would scratch the same itch. Writing a poker bot sounds hilarious and I bet a really good one could beat even the best humans but I'm not sure that's what I'm looking for. mit has a pokerbot thing each year, ive been trying to get good enough at java to write something for it (oh and work)
|
# ? Dec 15, 2015 23:55 |
|
quote:Thomas and I were attempting to attach a profiler to our Ruby code this morning (via cowboy coding directly on one of the GM servers), which might have been the cause of that. This was approximately 12 hours earlier than the time stamp on this post.
|
# ? Dec 16, 2015 01:17 |
|
got going with spacemacs again. it's great but goddamn icannot disable smartparens fast enough a garbage mode for idiots
|
# ? Dec 16, 2015 02:14 |
|
has anyone suggested that the stockfighter snafu might be part of the game??
|
# ? Dec 16, 2015 02:15 |
|
so just got my mac not very long ago and i'm going wild on swift and react, migrating old projects. more terrible programmer questions from me. reactjs and more specifically react native users if any: what do you usually use as a backend for rendering views on the server? i'd like something easy and nashorn and c# aren't really that at all. this is killing me and making me want to give up on react altogether. swift question if anyone here uses it: networking. am i really recreating the wheel trying to do something like java's remote/unicast? is the language still that much in its infancy that no decent libraries for persistence and communication exist? i can't find good resources for either questions online. i'm coming from a c#/java background and trying to make something other than simple client side stuff on the iOS. the few answers i have found have been a major pain in the rear end so i just feel like i'm looking in the wrong places.
|
# ? Dec 16, 2015 02:38 |
|
pepito sanchez posted:reactjs and more specifically react native users if any: What's not easy about c#? I use ReactJS.net and it was a breeze to set up, with the bonus of having react render server-side content on page load
|
# ? Dec 16, 2015 02:44 |
|
MALE SHOEGAZE posted:got going with spacemacs again. it's great but goddamn icannot disable smartparens fast enough I told one of my coworkers about spacemacs and he loves it. it looks so drat cool every time I see it going on his screen and those colored parens are fabulous, what's wrong with them?
|
# ? Dec 16, 2015 03:17 |
|
Lone_Strider posted:What's not easy about c#? I use ReactJS.net and it was a breeze to set up, with the bonus of having react render server-side content on page load apologies for possibly starting another code-styling war, but is it just me who looks at poo poo like this: code:
and thinks "what the gently caress, that indentation makes no sense. The initialComments line is a logical child of "new", not of @Html. it should not be just randomly indented, it should be placed in such a way that you'll immediately see that "initialComments" are a property of the second argument to the function, something like: code:
code:
it's not just an aesthetic complaint either, when I look at JS code i often find myself having to squint at the smorgasbord of braces and parentheses exactly because that kind of indentation is totally useless as a comprehension help NihilCredo fucked around with this message at 03:30 on Dec 16, 2015 |
# ? Dec 16, 2015 03:28 |
|
NihilCredo posted:http://fsharp.org/guides/math-and-statistics/ might interest you thats out of date enough that it lists ilnumerics as having an open/free version which they havent in a few years so
|
# ? Dec 16, 2015 03:30 |
|
rrrrrrrrrrrt posted:fyi this isn't the case except maybe at limited forms of heads up poker. the university of alberta published their solution for heads up limit poker and their method applies equally well to multi player and no limit games. the cost to generate optimal strategies is high so bots may not exist yet, but we know how to make them once costs come down. it's not even unlikely that someone has already generated the strategy for a limited version of heads up no limit, the estimated costs are in the low millions of dollars
|
# ? Dec 16, 2015 03:31 |
|
Ochowie posted:Basic data analysis, regression analysis. Also, some interpolation methods and finite difference methods. math.net numerics has at least some of that stuff. if there's a specific thing you wish existed there but doesn't lmk and maybe i'll contribute to an open sores lol
|
# ? Dec 16, 2015 03:31 |
|
NihilCredo posted:apologies for possibly starting another code-styling war, but is it just me who looks at poo poo like this:
|
# ? Dec 16, 2015 03:38 |
|
NihilCredo posted:snip I try to create my view models in such a way that I can just give react data directly and not whip up an anonymous object code:
also Shaggar posted:its because javascript is bad
|
# ? Dec 16, 2015 03:48 |
|
Lone_Strider posted:What's not easy about c#? I use ReactJS.net and it was a breeze to set up, with the bonus of having react render server-side content on page load i played around with it, BUT only a bit. Correct me if i'm wrong but doesn't this require using asp.net mvc with react.web.mvc nuget? otherwise i can't use jsx in visual studio itself? that's what seemed like a pain in the rear end to me. i was hoping for more of a purely webapi solution so i could use it from any other program in any other language i might want to code.
|
# ? Dec 16, 2015 03:49 |
|
I don't get why you'd use react instead of knockout
|
# ? Dec 16, 2015 04:00 |
|
i havent used react or knockout but i think it's because reactive programming is really good?
|
# ? Dec 16, 2015 04:02 |
|
I'm not sure I totally understand, but if you're using visual studio 2015 or visual studio code, you've got support for jsx syntax built-in w/o any additional packages or addons. are you just looking to get some kind of webapi spun up so you can point your react code to endpoints to get data?
|
# ? Dec 16, 2015 04:04 |
|
the talent deficit posted:the university of alberta published their solution for heads up limit poker and their method applies equally well to multi player and no limit games. the cost to generate optimal strategies is high so bots may not exist yet, but we know how to make them once costs come down. it's not even unlikely that someone has already generated the strategy for a limited version of heads up no limit, the estimated costs are in the low millions of dollars limit poker sucks tho and I thought those solutions were all play-not-to-lose style? I have a buddy who was doing his PhD at UofA but he was working on Go.
|
# ? Dec 16, 2015 04:05 |
|
Symbolic Butt posted:I told one of my coworkers about spacemacs and he loves it. it looks so drat cool every time I see it going on his screen the colored parens are something like rainbow-delimiters-mode smartparens-mode is the one that when you press ( inserts () i usually don't mind it and sometimes it's useful but i can see how some people would find it infuriating
|
# ? Dec 16, 2015 04:06 |
|
Knockout's functioning through attributes basically defies good tooling and my experience with it was that it was a bitch to make a composable view model that applies to composable components. Maybe I was doing it wrong. I found react was nicer after using both of them although I guess I'd still consider knockout as it's a smaller script to make clients download.
|
# ? Dec 16, 2015 04:12 |
|
rrrrrrrrrrrt posted:limit poker sucks tho and I thought those solutions were all play-not-to-lose style? what a waste.
|
# ? Dec 16, 2015 04:20 |
|
Bloody posted:thats out of date enough that it lists ilnumerics as having an open/free version which they havent in a few years so Yeah I searched around for the free version before I realized that. On the plus side I wasted 15 minutes of work time trying to find it. Bloody posted:math.net numerics has at least some of that stuff. if there's a specific thing you wish existed there but doesn't lmk and maybe i'll contribute to an open sores lol I noticed it has a mono version. I might try it with ASP.NET 5 on linux. I know it's basically a tech demo but it should be enough for my lovely hobby apps.
|
# ? Dec 16, 2015 04:33 |
|
Symbolic Butt posted:I told one of my coworkers about spacemacs and he loves it. it looks so drat cool every time I see it going on his screen rainbow parens owns. smartparens tries to match delimiters for you automatically and maybe i just dont know how to use it but it blows i hate autoindenters/etc for whatever reason. i think they're meant for people who are more intentional about what they type DONT THREAD ON ME fucked around with this message at 04:51 on Dec 16, 2015 |
# ? Dec 16, 2015 04:45 |
|
Ochowie posted:Yeah I searched around for the free version before I realized that. On the plus side I wasted 15 minutes of work time trying to find it. parts of mono are probably pretty decent, math.net has been on it for A While. i think at least one primary dev works primarily in a mono environment, judging from their page
|
# ? Dec 16, 2015 05:04 |
|
MALE SHOEGAZE posted:rainbow parens owns. smartparens tries to match delimiters for you automatically and maybe i just dont know how to use it but it blows i like ctrl-k ctrl-d in visual studio or gg=G in vim
|
# ? Dec 16, 2015 05:05 |
|
i tried react a bit my initial impression is I would need to be forced to use it over knockout or even angular because it's not overtly better. reactive programming is a good idea but both knockout and angular have that capability if you build it right maybe using components and rendering explicitly is faster than just using $scope/observables and dom attributes? MeruFM fucked around with this message at 05:25 on Dec 16, 2015 |
# ? Dec 16, 2015 05:22 |
|
MALE SHOEGAZE posted:rainbow parens owns. smartparens tries to match delimiters for you automatically and maybe i just dont know how to use it but it blows yeah smart parents is good if you know how to use it because it's way more than just delimiter matching. it's got commands for basically everything you would do with an s-expr. it's really only useful in a lisp though.
|
# ? Dec 16, 2015 05:26 |
|
angular is bad and I haven't really used knockout so react wins by default
|
# ? Dec 16, 2015 05:51 |
|
Lone_Strider posted:I'm not sure I totally understand, but if you're using visual studio 2015 or visual studio code, you've got support for jsx syntax built-in w/o any additional packages or addons. are you just looking to get some kind of webapi spun up so you can point your react code to endpoints to get data? exactly it. i want this to work as well for a website (which is our initial semi-big project in ~three months) as well as an iOS app built on react native. it's why my question was so vaguely worded. at the same time i'd like to get the most out of react with server-side rendering. i'll just have to read up more on reactjs.net i guess? i'm very comfortable with working in c# knockout isn't really an answer for me here. i know it's your fav shaggar.
|
# ? Dec 16, 2015 05:54 |
|
rrrrrrrrrrrt posted:limit poker sucks tho and I thought those solutions were all play-not-to-lose style? play not to lose is the winning strategy in poker. if your play is unexploitable then every non-optimal move your opponent makes is more money for you. also the solution works for no limit, but they don't yet have the resources to actually simulate the problem space
|
# ? Dec 16, 2015 06:22 |
|
i hope the open bot tournament is razz or something razz is fun
|
# ? Dec 16, 2015 06:39 |
|
rrrrrrrrrrrt posted:limit poker sucks tho and I thought those solutions were all play-not-to-lose style?
|
# ? Dec 16, 2015 06:42 |
|
Luigi Thirty posted:i hope the open bot tournament is razz or something It's razz and the tournament is actually a Voight-Kampff test. If the bots rage quit they pass.
|
# ? Dec 16, 2015 07:26 |
|
|
# ? May 26, 2024 06:02 |
|
Ochowie posted:Is there a c# or java equivalent to numpy/scipy/pandas? no
|
# ? Dec 16, 2015 07:43 |