« first day (1105 days earlier)      last day (3376 days later) » 

8:16 AM
Am I the only one who think WRI new web site since V 10 is so annoying? I feel like I am visiting Yahoo each time I now go to WRI sites. With all the annoying pop-ups that keep resizing and adjusting and showing up as I move the mouse. I prefer the old simple website without all this flashy and way over done javascript decorations all over. Browsing the web is becoming no fun any more.
 
 
1 hour later…
9:36 AM
@Nasser I agree. It is faster for me to find something there via Google... not the best sign. But UserExperience department was running a survey about that so maybe something will change.
@Mr.Wizard Does this deserve separate question with tag? mathematica.stackexchange.com/a/72240/5478
 
 
3 hours later…
12:47 PM
@halirutan I see it as Mathematica now, but I don't know if it was your override that did the trick or if it picked it up on its own.
 
 
3 hours later…
3:48 PM
I can't believe I'm unable to find this (via Documentation, Google or SE): how do I get all (contiguous) subsequences of a given list? (Or substrings of a string.) It seems like one of the simplest combinatorial tasks, but I can't seem to find something like a built-in. (I know how I could do it in general, but I feel there should be a Mathematica-y way, and I just fail to search for the right keyword.)
 
@MartinBüttner ReplaceList?
 
awesome, thank you!
 
np :)
 
is there an equivalent to {a___,b___,c___} for strings instead of lists?
I tried ReplaceList["abcaa", a___ ~~ b___ ~~ c___ :> b] but it doesn't return anything
oh, StringReplaceList
ignore my monologue
 
4:14 PM
Any Linux users around?
I'm having trouble getting FileNameJoin to work
FileNameJoin[{"/path/with spaces", "file.txt"}, OperatingSystem -> "Unix"]
should return "/path/with\ spaces/file.txt", right?
 
okay, figured out my substring/sublist problem. would it make sense to post that as a question and self-answer it, seeing that google doesn't currently turn up any results for it?
 
@MartinBüttner Yeah, that sounds reasonable.
 
4:50 PM
done :)
 
Hey everyone, I've got a function hess[x] which seems to be disproportionately expensive
It looks something like `hess[x_] :=
Table[If[k == l, Sum[3 (x[[k]] - x[[j]])^2 - a[[k, j]]^2, {j, n}],
0.0] - (3 (x[[k]] - x[[l]])^2 - a[[k, l]]^2), {k, n}, {l, n}]`
a is an n*n matrix, and x is a n-vector.
For n = 300, this takes about 9 seconds, which feels too long to me
Is there anything I'm missing that could speed this up?
 
5:09 PM
Cook your own: LFNJ = StringReplace[
FileNameJoin[#1, OperatingSystem ->
"Unix"], " " -> "\\ "] & ;
LFNJ[{"/path/with spaces",
"file.txt"}]
 
 
2 hours later…
7:14 PM
0
Q: Why are many answers posted as comments?

John McGeeI have noted that many questions appear to have 0 answers, when there are several answers given as comments. What is the exact protocol for when an answer should be posted as an answer (or not)?

 
7:50 PM
@DavidZhang With[{diffsq = 3. Outer[Subtract, x, x]^2 - a^2}, IdentityMatrix[n] Total[Transpose@diffsq] - diffsq]
With[{diffsq = 3. Outer[Subtract, x, x]^2 - a^2}, DiagonalMatrix[Total[Transpose@diffsq]] - diffsq]
 
8:21 PM
@MichaelE2 Wow, that's a few orders of magnitude improvement! So the trick is to always prefer vectorized operations?
 
Not sure if anyone uses the WPC here, but this doesn't seem to work `Show[ExampleData[{"TestImage", "Lena"}],
Epilog -> {Red, PointSize@.05, Point[{256, 256}]}]`. I'd like to know why (although I suspect the answer is: bug) or perhaps learn if someone has been working with WPC and has come up with some workarounds.
Replacing Lena with Graphics[Disk[{256, 256}, 256]] and the Epilog is drawn.
 
 
2 hours later…
10:03 PM
@rm-rf I totally forgot to tell you: I have been in contact with the developers in the mean-time. It was simply a non-existing differentiation between various .m extensions (Objective-C, MATLAB, M, Mathematica).
See my comment here or read the full issue:
Additionally, I forked the repository and fixed the missing extensions .wl and .wlt. Please see the discussion of the pull-request. Unfortunately, there are a lot .mt extensions (really really many, compared to the few Mathematica tests on GitHub) so that we decided to leave
this out until a proper heuristic for all of them is found.
In summary: As soon as the current code base goes live on GitHub, all .m files that contain some comment at the beginning of the file will be successfully recognized as Mathematica.
Additionally, if you want to fix your language statistic on GitHub for Mathematica files, just create a .gitattributes file in the top-directory of your repository with the following content:
*.m linguist-language=Mathematica
*.wlt linguist-language=Mathematica
Be aware that this only fixes the language statistic, not the syntax-highlighting!
 
10:45 PM
@bobthechemist I'm not sure whether I would consider this a bug. I would consider it a feature that it works in the FE, because Show is for combining graphics, not images.
Representing an image as graphics raster, like in the good old times, instantly makes it work on the WPC:
@bobthechemist I haven't found any hints in the doc to Show that it can be used with Image.
 
Have any one used the ClusterIntegration package?
 
11:16 PM
Non-transitive Grime Dice
http://community.wolfram.com/groups/-/m/t/428745
 

« first day (1105 days earlier)      last day (3376 days later) »