« first day (73 days earlier)      last day (4412 days later) » 

1:51 AM
2
Q: Do we need the "expressions" tag?

VerbeiaThere are 11 questions tagged expressions including one that is tagged only with that tag. I wonder if this is too general a tag to be useful. What do others think? And how should we tag this question otherwise?

 
 
2 hours later…
3:42 AM
@Verbeia And, tomorrow we'll hit 1000 questions. Not bad at all for 73 days in beta.
To reiterate: no other site our age has anywhere near our stats, period. Sorting by newest on StackExchange makes this abundantly clear. (Although, Computer Science may be giving us a run for our money.)
It takes until Travel to start to exceed our stats, and they've been in beta for 283 days! We just lack users, and those numbers are going up constantly. I think we're in phenomenal shape.
Lastly, Skeptics just graduated with a little over a year under their belt. They trounce us on visits per day (4.2k v. 1k) and users (5.8k v. 883), but their question and answer count isn't out of our reach in another two months. I think we'll hit it before then.
I don't see that we yet have the user base to be considered for graduation at 90 days. But, I wouldn't be surprised if it isn't much beyond that.
 
CHM
4:55 AM
@rcollyer nice analysis. I think so too.
 
 
5 hours later…
9:44 AM
Hello
 
@Rojo I think I see how that works, and it's bloody brilliant, but I want to hear your own explanation, please.
 
I can explain how the first thing works well, but the speedup is more luck than brilliance
What were you referring to?
I honestly wasn't expecting such speedup
 
First thing? I'm confused now. I'm talking about the Block trick.
 
I mean, I suggested a change in your f1 that might cause doubt as to how it works. The change I suggested in your f2 was just adding a block for speedup. And I am still thinking why it speeds it up soo much
 
I missed the change (besides Block) to f1.
 
9:50 AM
Well, the Block just came to me as a means to remove the {{q__}}:>{q} as you asked
by preventing the recursion until after flattening
Then I saw a tremendous speedup
and started playing with it
 
I see. Very interesting. This is probably some kind of pattern matching "tail recursion" -- Leonid would be the one to know.
 
Yeah...
 
Even if it was luck you're still brilliant. :-)
 
Thanks, thanks... My luck has always been brilliant :)
I'm really interested in knowing why it helps SO much
I hadn't figured that there was a bottleneck in your code
that this somehow prevents
 
One minute...
Okay, read this. I once thought I understood it but I've since lost my grasp of it.
8
A: Tail call optimization in Mathematica?

Leonid ShifrinI can summarize the conclusions I was led to by my personal experience, with a disclaimer that what follows may not be the entirely right explanation. The anwer seems to lie in the differences between Mathematica call stack and traditional call stacks, which originates from Mathematica pattern-de...

 
9:55 AM
Let's see
 
Then, tell me if you think it has any relation to this case.
 
Humm
 
I just updated my answer; please check that it is as you meant.
 
I'm trying to make it fit as an explanation but I can't see how
Haha, thanks
Let's do some tests
 
I think I'm barking up the wrong tree with that tail call optimization thing. I think this is a simpler situation. The function is essentially Blocked throughout the first tree direction, so evaluation (with Block) operates in reverse. Am I making sense so far or talking gibberish?
 
10:09 AM
Why reverse?
It goes from the outermost f1 to the innermost f1 because the innermost don't exist until the outermost gets evaluated
It's just that
StringReplaceList doesn't get nested
Right?
I'll play to see if nested rules is a problem in general
 
This is apparently not the best time to be working on this; my brain refuses to go into first gear. :o)
 
Hehe, ok
 
Basically if I'm right your method prevents multiple tree traversals.
 
Why?
 
@Rojo sorry, I was looking at the code. I started to reason through it but like I said, I just cannot think right now. Do ask Leonid if you see him, will you?
 
10:16 AM
Definately
You let me know if you have a good theory too, same here
 
F'x
@MrWizard thanks for the time you invested in this string manipulation question
I’m afraid I might need more that one lifetime to understand all the tricks and niceties in all of the answers, especially yours and RM’s, though
 
 
4 hours later…
2:01 PM
Understanding the Grid style specifications requires a Ph.D in cryptography.
 
F'x
@SjoerdCdeVries maybe you could do with a PhD in decryption, which I understand is easier than cryptography as a whole :)
 
@F'x It took me 10 minutes to figure out how get a black background in the first row and column and alternating grey/white rows in the rest of the table.
 
F'x
@SjoerdCdeVries I looked at your linkedin profile yesterday and saw that it displayed you as a "3rd degree" connection
I am always amazed at this "6 levels of separation" kind of stuff :)
 
2:16 PM
@Fx It's probably via one of our Mathematica contacts. Or not?
 
F'x
@SjoerdCdeVries no, a college friend who worked at Thales for some time
I don't have any Mathematica contacts, I only started using the darn thing a few months ago
now I'm hooked :)
 
I know who you're talking about
@F'x Vincent C?
 
F'x
@SjoerdCdeVries yes
 
Try to replicate this...
 
F'x
@SjoerdCdeVries well, I hope you're not planning on putting this into a powerpoint slide, next to a bullet point saying "we clearly see that …" :)
@SjoerdCdeVries I won't try to, I avoid MMA for styling as much as possible
 
2:21 PM
@Fx Just for printing purposes
 
F'x
@SjoerdCdeVries good
you have some inconsistency in significant digits displayed, though
 
I know, but I don't care at the moment
 
F'x
@SjoerdCdeVries I'm sorry, after a few years of teaching it has somehow become a reflex
 
@Fx teaching what?
 
F'x
@SjoerdCdeVries theoretical chemistry, programming, statistical physics, that sort of stuff
 
2:24 PM
All numbers were rounded to two places, but if it results in .00 MMA skips them
 
F'x
but I'm always amazed that many of the young students who have a engineering degree (or sometimes a PhD) still don't manage to care about that kind of stuff
anyway, I'm ranting, should stop
 
If I really want to display 2 digits I'd had to use NumberForm or so.
 
F'x
@SjoerdCdeVries by the way, I know I flag a lot of comments as "obsolete" because I think the hardest thing for a moderator is to keep these clean (and users often don't delete their obsolete comments)
 
@SjoerdCdeVries something like Grid[table, Frame -> All, Background -> {None, {None, {Gray, None}}, {{{1, 1}, {1, -1}} -> Black, 1 -> Black}}]
?
 
F'x
really feel free to reject them, or drop me a line if it's too much to your liking
 
2:29 PM
@Heike I used `Background -> {{{}}, {{White, LightGray}}, {1 ->
Black, {{1, 1}, {1, 21}} -> Black}},`
 
ah, yes, my construction for the alternating rows could have been simpler
 
which is almost the same. But it's really a puzzle.
 
I agree
 
-1 is better than 21
 
Colouring a row while preventing blending seems overly complicated.
Oh, wait, this also works: Grid[tab, Frame -> All,
Background -> {None, {Black, {Gray, None}}, {1 -> Black}}]
 
2:34 PM
@F'x I feel they are helpful. It might be that once in a while I handle them inappropriately when I'm working on my iPad. There are two flag screens, and both have a delete and dismiss link, but in one this refers to the question and in the other to the flag and on the iPad I don't get a tooltip and it's easy to make a mistake there.
 
I originally thought you wanted alternating columns for which I needed the other construction.
 
@Heike This is simpler indeed, moving the exception line into the row definitions
 
F'x
ok, gotta go, see you later
 
or is it the column definitions?
 
bye
 
2:36 PM
bye
 
@SjoerdCdeVries The second argument is for rows I think.
 
@Heike Yeah
Have to go to the AH. CYa
 
On Saturday afternoon? good luck with that.
 
 
6 hours later…
F'x
8:54 PM
p = ParametricPlot[{Cos[t] - Sin[t]^2/Sqrt[2], Cos[t]*Sin[t]}, {t, 0, 2 \[Pi]}];
l = Cases[Normal@p, _Line, Infinity][[1, 1]];
Graphics[{Blue, Polygon@l}]
 
9:05 PM
We are 999 questions. Annoyingly, I can't think of something to ask right now. Surely someone wants the warm feeling of getting us to 1000?
 
9:37 PM
@verbeia DOn't know a good one either. What about the Notebook history menu item? Or one about a repository of good stylesheets?
 
9:59 PM
1
Q: What criteria should guide us to flag new questions on SO for migration here?

VerbeiaThere are quite a few new questions tagged Mathematica on StackOverflow. I see us as potentially missing out on some new users. How should we decide which questions should be flagged to be migrated? This question is separate from this question, which was mainly about old questions.

 
acl
10:24 PM
brilliant: I have a problem with NDSolve, probably trivial and caused by me staring at the same code for the whole day. I decided to simplify the example down so as to ask here. The slightest, most trivial modification to the code fixes the problem...
Of course, I really do need to solve the unmodified problem, which I can't. excellent
 
F'x
10:40 PM
I don't really understand why that question would need being protected?
4
Q: Solve Lagrange multipliers optimization problem

Martin SchererI have two nested solid figure, where $V(a,h,\tau)$ defines the volume and $A(a,h,t)$ defines the surface. The outer solid figure is parametrized in $a_s$,$h_s$ and $t_s$ (they share a common center). Now I raise the Volume of the inner by factor $\alpha$, so $$V(a_s,h_s,t_s) = \alpha V(a,h,t)$...

 
@F'x which one?
 
F'x
the one I just gave the link to
it's not one of those funny or exciting everyone-will-want-to-add-their-comment-on-it questions
 
Sorry I was too fast -the fact that it got three answers that were deleted either by the owner or the moderator suggests RM wanted to make sure it didn't attract any drive by answers.
 
F'x
@Verbeia hum, I'm not high-rep enough to see deleted answers (god, it feels bad to be a non-mod!!)
@Verbeia thanks for the information
 
11:10 PM
1000 questions
Nice
 

« first day (73 days earlier)      last day (4412 days later) »