the algorithm finally works ….

It has been ages…trying to get the algorithm works ( alan suggested this and yet  everything need to be done from scratch!!!) Dealing with player/stage is another challenging task and to understand it is another story. Once this algorithm works then it can be enhanced and hoping that it will works… The thesis seminar is rescheduled and I think I need few more weeks to run the algorithm and code is only half way though…The anchoring should not occur if there is 1 or 2 failing robots and my results show the other way around…..and after 2 months, I finally get the exact result ( I think ) …plotting with matlab is a definite easy this time  as I have the m-code written from the previous exercise and I just change the data ( thanks to matlab) ..  and box plots are ready for analysis ( still part of it ). Now I realise why susan really encourages us on using Matlab for analysis..looks how clean the graphs are and most importantly, you can save the output to pdf (this is needed as I am writing with latex ) and it is clear enough even how small you resize the diagram.

Current issues that I have is the charging part…I also need to have a rigorous testing with the algorithm that I proposed. we might need to increase the number of robots, increase the number of failures and lots more…to see whether the first and second algorithm can really solve it, We might need to have more sophisticated way of solving the issues according to granuloma formation. While preparing the movie for discussion and presentation…I am trying to add energy to the robots which is fine but..the charging issues need to be solved by this week or else…I can’t show to Jon during meeting.

End of this month, we need to attend the conference in edinburgh and I will be visiting edinburgh and glasgow…hopefully the codes work before the trip!!

Here are some of the movies that I captured while I am running the experiment. With one and two failing robots injected, the swarm still can arrive to the beacon. But once more failures are injected, the stagnation starts to occur…

Advertisements

LaTeX Poster Template

Do you have any latex poster template to share?  I am preparing a poster paper for a workshop this month.  I couldn’t find any poster templates in LaTeX that I liked, and frankly I have started preparing it using open office instead. However, just wanna give it a try preparing it using latex. Found the following template form Mr. google and I think it suits my poster. I prefer to produce a `simple’ but `elegant’ type of poster and since this paper have lots of graphics rather than result, I think I would like to only have two-column instead of three.

I should remember that preparing a poster is very different from preparing a paper. You will not be enlarging your research paper and wallpapering the display board. Your main objective in preparing text for this presentation is to edit it down to very concise language. Use bullets and numbers to break text visually and aid you in the interactive use of your poster.

This is how the poster looks like ( p/s: still have way though….)..



The final version: to be updated after tomorrow meeting ok?

Meeting outcome: finalised paper to be submitted and Alan is happy to involve and go though the paper..:P..The poster needs more graphic…means another version need to be prepared…and finally finally…ICARIS deadline and all experiment should ready this week to be discussed with Jon next week.

Using Multiple X- and Y-Axes in Matlab:

I am trying to plot another graph having Double Axis Graphs for my simulation.  lets see whether it works or not. This is the output…So the analysis shall starts…

Next meeting: Wedenesday 1 p.m over in electronics

Basic Curve fitting in Matlab

t = 0:2000:20000;

p=[10.04  6.02  5.16  3.6 5  4.71  4 2.93  1.51  1.82  0.03 ]

figure1 = figure;

axes1 = axes(‘Parent’,figure1);

box(axes1,’on’);

hold(axes1,’all’);

a= plot(t,p,’o’);

axesLimits1 = xlim(axes1);

xplot1 = linspace(axesLimits1(1), axesLimits1(2));

yplot1 = ppval(fitResults1, xplot1);

fitLine1 = plot(xplot1,yplot1,’DisplayName’,’   shape-preserving’,…
‘Tag’,’shape-preserving’,…
‘Color’,[0 0.5 0]);

Thoughts and Ideas on plotting graph in Matlab

Had a meeting with Jon via skype today as the weather was bad. Discussed on some experiment that I should have before 4th Jan. Finally managed to plot multiple error bar in matlab ..lots more work !!!

load nofault.dat;
load fault.dat;
a = mean(nofault, 2);
y=std(nofault,1,2);
z= y/sqrt(10);
b = mean(fault, 2);
c=std(fault,1,2);
d= c/sqrt(10);
figure;
h1=errorbar(a,z);
hold on;
h2=errorbar(b,d,’–‘);

Output:

excel/gnuplot/matlab?

Sedang analysis result nih, its good if I can decide which method to use in analysing my data…started with excel…susah juga nak multi- lines!! spend a few hours..and malas dah nak fikir sebab memang susah sangat la…lagi2 kalau nak add error bar and stuff, change to gnuplot..tengah syok2 belajar dengan lau..tetiba datang la my friend Mark saying better use matlab which is much more simple and you can analyse your graph in many2 ways…so let give it a try..I simply plotted the bahaviour 4 robots…memang result dummy sangat tapi better try it yer…

load robot.dat;
x = robot(:, 1);
y = robot(:, 2);
z = robot(:, 3);
a = robot(:, 4);
plot(x, y, x, z, x, a, ‘-r’, x , y, x, z, x, a, ‘ok’);
the resu;t is voila:
Bagaimana? I took 1 hour to draw this using GNUPLOT….jadi esok boleh la ke lib and pinjam buku matlab pulak…:P
Selepas menambah sedikit cosmetic….jadi la graph itu begini, which is not difficult and agak mudah nak tambah cosmetic ini ya..:)
Nasib baik la mamat nih bgtau suruh pakai matlab, or else seharian la dengan gnuplot yang susah nak generate pdf file dia…voila…
dan seterusnya sedikit result untuk di discuss esok..malam ini boleh rehat2 skit ye…blurgh!!!