From: Igor on 8 Aug 2010 14:01 is this 'as it should be' ? If yes, what it can be useful for? =========================== classdef a_class %A_CLASS Summary of this class goes here % Detailed explanation goes here properties p end methods end end =========================== N=3; for i=1:N a=a_class(); a.p=single(i); A(i)=a; end B=zeros(1,N); B(1:N)=A(1:N).p; %same as B(1:N)=A(1).p (???) B %returns 1 1 1 ... A(:).p %returns 1 2 3 ... (yet,this is neither an array,nor cell array ) ============================ I've expected something like the result of B(1:N)=cell2mat({A(1:N).p}); is there any better way,then " B(1:N)=cell2mat({A(1:N).p}); " ?
From: Igor on 9 Aug 2010 04:55 Yep... In fact,the thing was [A(:).p] was generating error, when it's related to AccelerEyes' Jacket. (A.p is gsingle ). Seems like it's their error. My fault was not to test that on usual matlab... Thanks for your responses.
From: Igor on 9 Aug 2010 06:21 "Bruno Luong" <b.luong(a)fogale.findmycountry> wrote in message > > Might be you could try cat(1,a(:).p) (with gsingle type) to see if it works. > > Bruno Thanks for suggestion... Well..Not working either.... ==================================== K>> B=cat(1,ALL_m2(:).alpha) ??? Error using ==> cat internal error: too many internal allocations (calls.cu:327) ==================================== well... for now I've simply replaced that with ==================================== for i=1:numel(ALL_m2) ALL_angles(i)=ALL_m2(i).alpha; end ==================================== That's not a performance bottleneck , in fact.... I'll report that behavior to them, may be they'll fix that...
From: Igor on 9 Aug 2010 09:10 "us " <us(a)neurol.unizh.ch> wrote in message > - the error message indicates that your hw does not know how to deal with > big arrays... > - unfortunately, we don't have the GPU (or sw) to reproduce this error... > > us Yep...thanks... As I've mentioned before ... > Yep... In fact,the thing was [A(:).p] was generating error, when it's related to AccelerEyes' Jacket. (A.p is gsingle ). Seems like it's their error... > I'll report that behavior to them, may be they'll fix that...
|
Pages: 1 Prev: Is it possible to vectorize a for loop with if statements? Next: solving symbolic equation |