@Öskå There's another cluster elsewhere which does, so people who use openMPI use that. That leaves more power for the rest of us, which is nice. And it seems most who do heavy calculations do use MPI, so I can use as many cores as I want, which is also nice!
@Öskå I see, well that's more or less what I tried too. Decided I'd only do it if I really had to. I think I did use some code using openMPI eventually, but not written by myself
too complicated, I already have trouble understanding the physical problems I'm working on. Adding programming complexity confuses me (which isn't hard though)
@seismatica I also didn't have a keyboard, mouse or screen. I didn't manage to get NOOBS working. NOOBS won't auto-acquire an IP address and have ssh set up by default, but Raspbian will. So you can put Raspbian on your SD card (google how), then figure out the IP address of the Pi somehow and just connect with ssh.
@Szabolcs does a Raspbian installation the same as a NOOBS installation? Will I get MMA from that? Would you mind sharing any link about installing Raspbian on the Raspberry using a laptop's keyboard and mouse? I don't want to follow the wrong method and get stuck. I might also bite the bullet and buy a screen for the Raspberry: what do you think about this LCD screen: amazon.com/Digital-Reversing-Satellite-Receiver-Equipment/dp/…?
@Pickett Why? According to the docs, shouldn't it rotate the cube about the axis passing through the line connecting $(1,1,0)$ and $(2,1,0)$?
Which is what it actually does?
So, like you take the cube in your first picture and push it over backwards (towards the infinite y direction) until it topples. Once it topples it looks like your second plot
Or am I missing something? I never used these functions
I think NOOBS includes several OS and lets you select which one you want to use. But you need a screen/keyboard to do that. Or you can just install a Raspbian image on the SD card and it comes with ssh already set up.
Should I assume that nobody else mamaged to reproduce this behaviour ?
Background:
Mac OSX 10.9.4, Mathematica 9.0.1 vs. Mathematica 10.0 (both versions are Student Editions).
I have a notebook. It used to evaluate fine in Mathematica 9. I upgraded to 10 and, without doing any modification at a...
The curious thing is that a definition on a parallel kernel affects evaluation on the main kernel, even though the definition is cleared from the main kernel and the evaluation is not parallelized.
Is there a way to programmatically get a list of modules loaded by MMA (9 in my case) by default? I.e., the list should NOT include the list of modules that I have placed within init.m. If this is a complicated question, I'm happy to type up a question about it.
This is basically awesome... What Carl Jung, a Swiss psychologist called "Synchronicity"... Two independent simultaneous posts on same topic: http://community.wolfram.com/groups/-/m/t/326240 http://community.wolfram.com/groups/-/m/t/325956