« first day (837 days earlier)      last day (3945 days later) » 

04:41
Hi, I have a question about exporting graphics. I usually use the hpc on our university to do calculations, and the workflow is like this: first I request an interaction job on the hpc, and then I lunch a kernel on the interactive node I get, and connect to my frontend on my local computer.
This works great except sometime it give me problem when I need to generate a move or many graphics. I learned that the graphics are always rendered at the frontend, which slows down a lot of the generating of graphics or movie, since my local laptop is quiet slow.
Is there a way to export the movie or graphics without invoking my slow local frontend?
Is there a way to start a frontend at the hpc, or even many frontend so that I can export the graphics in parallel?
 
13 hours later…
17:31
@xslittlegrass maybe you can use Rasterize[#,"Image"]& and export the result (as .jpg or something)?
@chris hello
18:26
@JacobAkkerboom I guess the slow part is that the rasterizing process is completed on my local computer
18:46
@xslittlegrass yes: you need to specify something like
SetOptions[Plot,DisplayFunction->Identity]
a=Plot[Cos[2*x],{x,0,Pi}]
Export["a.eps",a]
Exit[]
@chris I tried this, but always get the error like "can't open display 0:0"
@chris I guess that also uses the frontend to render the graphics, but just in the background. The problem is that there seem to be no x11 on our hpc so that mathematica can't start a fontend
@halirutan on your recent graph answer if we get a 'branch' with, say, relations r [ g[b] , b[g] ] how can I get (using a generic approach) each sub-branch r[g[b]] and r[b[g]] ? I tried various Maps and Applys
for clarity, an ideal output would be { r[g[b]] , r[b[g]] }
19:25
@xslittlegrass it works for me without x11. are you sure you don't have something in your init.m file?
19:43
@chris thanks for letting me know, let me double check
@chris I have nothing in my init.m
@chris Also there it also says on that page you linked "the following is for Mathematica < version 6; for higher versions you shouldn't expect graphics to work without a windowing system"
@chris what's you version? I'm using version 9
20:40
@PlaysDice
paths[x_?AtomQ] := x
paths[x_] := Head@x /@ Flatten[paths /@ List @@ x]
paths[a[b, c[d[f, g, h], e]]]
{a[b], a[c[d[f]]], a[c[d[g]]], a[c[d[h]]], a[c[e]]}
21:04
@MichaelHale. Thank you! That's what I wanted.
 
2 hours later…
22:59
Could someone running version 9.0.1 please confirm this reported problem with my code:
The first time I run this, the tag appears twice: When i = 0 When i = 0 Power::indet: "Indeterminate expression 0^0 encountered."cyberSingularity 42 mins ago
@Mr.Wizard I got the tag twice too
@Rojo Do you have time and interest to chase down the problem? I do not have it in version 7.
23:15
@Mr.Wizard When I run your code (first 3 codeblocks), then I get every message only once.
(Linux, V8 and V9.0.1)
@halirutan Very interesting. @Rojo please remind me, what OS do you use?
23:31
@Mr.Wizard Windoes
8, x64
Thanks.
Humm, Probably I won't be tracking it.
I see that I had looked into that question at that time, but I forgot all
and, it's not that bad if all that happens is that the first time you see it twice, right?
@Rojo I suppose not. Still, it's a bit strange, and you were the one to fix an earlier version-related bug in it. Nevertheless I understand that there are more interesting things to work on!
23:35
@Mr.Wizard I'll take a look now and see if some fixes comes up more or less easily
OK :-)
23:52
@Mr.Wizard Are you sure it isn't the same issue as before?
@Rojo I thought that bug was corrected; why would it appear again with this code? (Sorry if I have wasted your time.)
Hehe
I mean, I think the problem before was the same
and for some reason I had suggested putting Identity@ somewhere
and then it was refactored 3453 times
@Rojo Let me look at the edit history...
23:58
@Mr.Wizard Yeah
@Rojo Well nuts, I thought I ported your fix but I guess I never did. I am sorry. Would you please try to fix the current version, assuming it is not difficult?

« first day (837 days earlier)      last day (3945 days later) »