MATLAB: How to plot multiple data sets from different files all on the same plot

plot

I'm trying to write a script that does the following: Reads in data from a file _aoi_aware to see which aoi is in use (i.e. which data it needs to plot)

Then prompts the user to open a file based on the determined aoi

Then plots the data in the aoi file, which is in the form of coordinates.

The only problem with this is that the original file that contains which aoi is in use can contain multiple aoi's. I will need all of these plotted. So ultimately, I would like the user to be prompted to load the aoi_aware file, then be prompted to select data from aoi_1, if that is open, and then have the script plot that. But if aoi_2 appears later in the aoi_aware file,then that needs to be plotted as well. So I want the script to check for all aoi's in use and plot all of them on the same plot. Here's the code I have so far, and it keeps looping on aoi_4 and producing errors. I'm not sure how to correct the code I've written to achieve the desired result. All help is appreciated. let me know if I can clarify the question.

[fileName3 pathName3]= uigetfile('~/*_aoi_aware_out_Z.txt','Select the appropriate test file under *_aoi_aware_out_Z.txt');
filepath3=fullfile(pathName3,fileName3);
%%Read columns of data according to format string.
[TimeHdr CameraID_1 CameraID_2 CameraID_3 CameraId_4 CameraID_5 CameraID_6 CameraID_7 CameraID_8 Track_aoi Track_camera]  = textread(filepath3, repmat('%s ', 1,11),1);
[Timecell Cam1 Cam2 Cam3 Cam4 Cam5 Cam6 Cam7 Cam8 AOI Cam]=textread(filepath3, '%s %f %f %f %f %f %f %f %f %f %f', 'headerlines', 1);
UTtimeCharArray = char(Timecell);
years = str2num(UTtimeCharArray(:,1:4));
months = str2num(UTtimeCharArray(:,6:7));
days = str2num(UTtimeCharArray(:,9:10));
hours = str2num(UTtimeCharArray(:,12:13));
minutes = str2num(UTtimeCharArray(:,15:16));
seconds = str2num(UTtimeCharArray(:,18:21));
times = datenum(years,months,days,hours,minutes,seconds);
%%Check to see which AOI it's tracking on each min.
for i=1:length(times)
    if AOI(i)==1
          filepath_aoi1=fullfile(pathname_aoi_1,filename_aoi_1);
          [TimeHdr CameraID1 Col1 Row1 Quality1 Width1 RA1 Dec1 Error_est1]  = textread(filepath_aoi1, repmat('%s ', 1,9),1);
          [Timecell cam1 col1 row1 qual1 wid1 ra1 dec1 error1]=textread(filepath_aoi1, '%s %f %f %f %f %f %f %f %f ', 'headerlines', 1);
          %%Plot



          figure(1)
          plot1=plot(col1, row1, 'c');
          grid on
          legend(plot1, 'row','column','Location','Best');
          datetick('x','HH:MM:SS', 'keeplimits')
          title(['TA Gyroscopic Drift' UTtimeCharArray(1,12:19) ' --> ' UTtimeCharArray(end,12:19) ' UT '],'FontSize',10)
          %xlim([min(times), max(times)])



          xlabel('column');
          ylabel('row');
          hold on
      elseif AOI(i)==2
          [filename_aoi_2 pathname_aoi_2]= uigetfile('~/*_AOI_2_out_Z.txt','Select the appropriate test file under **_AOI_2_out_Z.txt');
          filepath_aoi2=fullfile(pathname_aoi_2,filename_aoi_2);
          [TimeHdr CameraID2 Col2 Row2 Quality2 Width2 RA2 Dec2 Error_est2]  = textread(filepath_aoi2, repmat('%s ', 1,9),1);
          [Timecell cam2 col2 row2 qual2 wid2 ra2 dec2 error2]=textread(filepath_aoi2, '%s %f %f %f %f %f %f %f %f ', 'headerlines', 1);
          %%Plot
          figure(1)
          plot1=plot(col2, row2,'k');
          grid on
          %legend(plot1, 'column','row','Location','Best');

          datetick('x','HH:MM:SS', 'keeplimits')
          title(['TA Gyroscopic Drift' UTtimeCharArray(1,12:19) ' --> ' UTtimeCharArray(end,12:19) ' UT '],'FontSize',10)
          %xlim([min(times), max(times)])
          xlabel('rows');
          ylabel('columns');
          hold on
      elseif AOI(i)==3
          [filename_aoi_3 pathname_aoi_3]= uigetfile('~/*_AOI_3_out_Z.txt','Select the appropriate test file under **_AOI_3_out_Z.txt');
          filepath_aoi3=fullfile(pathname_aoi_3,filename_aoi_3);
          [TimeHdr CameraID3 Col3 Row3 Quality3 Width3 RA3 Dec3 Error_est3]  = textread(filepath_aoi3, repmat('%s ', 1,9),1);
          [Timecell cam3 col3 row3 qual3 wid3 ra3 dec3 error3]=textread(filepath_aoi3, '%s %f %f %f %f %f %f %f %f ', 'headerlines', 1);
          %%Plot
          figure(1)
          plot1=plot(col3, row3,'g');
          grid on
          %legend(plot1, 'column','row','Location','Best');
          datetick('x','HH:MM:SS', 'keeplimits')
          title(['TA Gyroscopic Drift' UTtimeCharArray(1,12:19) ' --> ' UTtimeCharArray(end,12:19) ' UT '],'FontSize',10)
          %xlim([min(times), max(times)])
          xlabel('column');
          ylabel('row');
          hold on
      else AOI(i)==4
          [filename_aoi_4 pathname_aoi_4]= uigetfile('~/*_AOI_4_out_Z.txt','Select the appropriate test file under *_AOI_4_out_Z.txt');
          filepath_aoi4=fullfile(pathname_aoi_4,filename_aoi_4);
          [TimeHdr CameraID4 Col4 Row4 Quality4 Width4 RA4 Dec4 Error_est4]  = textread(filepath_aoi4, repmat('%s ', 1,9),1);
          [Timecell cam4 col4 row4 qual4 wid4 ra4 dec4 error4]=textread(filepath_aoi4, '%s %f %f %f %f %f %f %f %f ', 'headerlines', 1);
          %%Plot
          figure(1)
          plot1=plot(col4, row4, 'b');
          grid on
          %legend(plot1, 'row vs. col','Location','Best');
          datetick('x','HH:MM:SS', 'keeplimits')
          title(['TA Gyroscopic Drift' UTtimeCharArray(1,12:19) ' --> ' UTtimeCharArray(end,12:19) ' UT '],'FontSize',10)
          %xlim([min(times), max(times)])
          xlabel('column');
          ylabel('row');
          hold on       
      end 
      i=i+1;
  end

Best Answer

...
for i=1:length(times)
  if AOI(i)==1
    ...
  elseif AOI(i)==2
    ...
  elseif AOI(i)==3
    ...
  else AOI(i)==4
    ...

The problem about continuing on case 4 is there's no 'if' on the 'else' so the AOI(i)==4 is meaningless--I'm somewhat surprised Matlab didn't flag it but didn't look at the code in the editor to see if it did give any warning.

If there are only four possible AOI values 1-4, then fixing that to also be elseif will stop that; if there are additional AOI values possible then unless you want them ignored you need to deal with them somewhere. Can't answer that as there's not info on that provided.

If there's a range of values after 3 that all have identical processing then it's likely the SWITCH construction might make writing the logic simpler than the nested IF

Related Solutions

MATLAB: How to create dynamically changing legend entries based on an unknown number of plots

I modified your practice code a little...it keeps the legends in order for those used for the SWITCH construct--there's not enough information given to know who to use for the added cases at the bottom but do the same thing...I presume you have some way to know who's who when you add the additional cases.

If it were me, I'd also throw the linestyles into an array and select from it and rearrange the data to use either named fields or cell arrays or something to eliminate the CASEs full of identical code other than some data values but that's not significant re: the ? on legends and keeping them consistent w/ the data. (I did do some more looking at it and there is no easy way to simply add an additional label to an existing legend that I could come up with anyway that was any more straightforward than the equivalent of this.)

You could keep an array of the legend strings and then just build an index vector to them to use to pass to the legend() call instead of building the string array as here. There are many ways to skin the same cat... :)

strvcat() is deprecated but I had a brain cramp and couldn't get the cell concatenation alone to work the way I wanted it to--I ended up w/ an array of cell arrays every time instead of a cell array of characters so I just reverted to old friends... :)

hplt=[];
lgnd={};
first=true;
for i=1:length(uAOI)
switch uAOI(i)
  case 1
    hplt=[hplt;plot(x1,y1,'b*')];
    lgnd=cellstr(strvcat(lgnd{:},'first'));
    legend(hplt,lgnd)
    if first, hold on, grid on, first=false; end
  case 2
    hplt=[hplt;plot(x2,y2,'r*')];
    lgnd=cellstr(strvcat(lgnd{:},'second'));
    legend(hplt,lgnd)
    if first, hold on, grid on, first=false; end
  case 3
    hplt=[hplt;plot(x3,y3,'k*')];
    lgnd=cellstr(strvcat(lgnd{:},'third'));
    legend(hplt,lgnd)
    if first, hold on, grid on, first=false; end
  case 4
    hplt=[hplt;plot(x4,y4,'c*')];
    lgnd=cellstr(strvcat(lgnd{:},'fourth'));
    legend(hplt,lgnd)
    if first, hold on, grid on, first=false; end
  end
end
% add some more
% No info on who to use for legends given...
%hplt(end+1)=plot(newx,newy);
% following will not be what is desired, though
%hleg(end+1)=legend(hplt,'first','second', 'third', 'fourth');

MATLAB: How to add an additional overall legend for additional data overlaid on a 3×1 subplot, which already has a legend for each of the 3 plots

I had an example much more like the following in mind...

x=1:10; y=rand(12,10);                       % make up some data to plot
l=cellstr(num2str([1:size(y,1)]'));          % and labels for them...
ho=subplot(3,1,1:3); po=get(ho,'position')   % To get the positions overall
po =
  0.1300    0.1100    0.7750    0.8150
delete(ho)                                   % don't want, just positions
for i=1:3                                    % plot some of y on each
  i1=3*i-2;i2=3*i;
  hs(i)=subplot(3,1,i);
  plot(x,y(i1:i2,:))
  legend(l(i1:i2))                           % label the ones...
end
for i=1:3                         % ok, now shrink the plots to make
  p=get(hs(i),'position');        % room for the added axes to hold 
  p(3)=0.675;                     % the super legend
  set(hs(i),'position',p)
end
hl=axes('position',[0.85 0.11 .105 0.815]);   % and that axes
hlin=plot(hl,x,y); set(hlin,'visible','off'); % can't legend w/o lines
set(hl,'visible','off')                       % don't want to see axis
legend(hl,l)                                  % and the global legends

That something like the effect you were looking for? (See how little actual was req'd to display something to look at, btw? I did get it into 20 lines! :) Try to pare your questions down to the essence instead of a whole application for somebody to try to wade through).

Obviously, I did some "magic numbers" of the sizes for the position; that's why I got the overall positions vector to begin with--I could look at those numbers and figure out a set of values to use to get the effect. You can use more sophisticated computational ways to generate them based on actual size and code there.

As always, "salt to suit..." :)

NB: I did NOT take the time to match the line styles; I just let them cycle automatically. If you control which style matches which line in the overall with those in each subplot then the legends will match, too.

Best Answer

Related Solutions

MATLAB: How to create dynamically changing legend entries based on an unknown number of plots

MATLAB: How to add an additional overall legend for additional data overlaid on a 3×1 subplot, which already has a legend for each of the 3 plots

Related Question