Problem with multiple if conditions....

Hello there:

I have a code composed of 4 if, this are the condition for the xlim of plots that follow the condition. There are 4 situations that i name a=1, a=2, a=3, a=4. Each satisfy certain characteristics of the 2x2 matrix l, which is a line formed by two points :

figure(2)

clf
subplot(2,1,1)
hold on

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%ANGULAR CONDITIONS in the cartesian plane for a line (l)

for i=min(shore):max(shore)    

     plot(dp(shore==i,3),dp(shore==i,4),'.','color',[1/i 1/i 1/i],'markersize',20)

     j(i,1)=i;
  end

%%%%%%%%%%%%%%%%%%%%%
if l(1,1)>l(1,2) && l(2,1)>l(2,2) %situation 1
a=1   

xlim([0 dkm]) %positive quadrant
end

%%%%%%%%%%%%%%%%%%%%%%%%%%%
if l(1,2)>=l(1,1) && l(2,1)>=l(2,2) %situation 2
a=2 

xlim([-dkm 0]) %negative quadrant
end

%%%%%%%%%%%%%%%%%%%%%%%%%%%
if l(1,2)>=l(1,1) && l(2,2)>=l(2,1) %situation 3
a=3

xlim([0 dkm]) %positive quadrant
end

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
if l(1,1)>=l(1,2) && l(2,2)>=l(2,1) %situation 4
a=4

xlim([-dkm 0])   

end
legend(num2str(j),'location','eastoutside')
errorbar(dp(:,3),dp(:,4),(dp(:,5)),'ok')%,'o','color',[1/10*i 1/10*i 1/i],'markersize',20)

box on
ylabel('elevation')
xlabel('along profile distance'

however, the if conditions doesn't work properly, in all the runs, only the a=3 condition is used, even if this not satisfy the condition. This is not logic. How can i do to write well this code?. Should i nest the if's?

any help or idea is appreciated... cheers

[Adam]

Use 'elseif' if only one should happen, although if they are mutually exclusive conditions it shouldn't make a difference. I guess they are not.

In your current case all you can tell from the final state of a being 3 is that the 4th condition never matched and the 3rd condition always does. It may match 1 or 2 also, but then matching 3 over-rides that.

Why does case 1 have > while all the other cases have >= ?

[Jules Ray]

Adam: only one case can occurr at time. Do you mena something like this:

if
elseif
elseif
else
end

???

[Adam]

Yep. That ensures mutual exclusivity at least.

[Jules Ray]

Thanks Adam, i did it in your way with an extra modification and it works now... the analysis using the l and its behavior is simply too weird, instead i used the difference between the elements of the matrix array and works perfectly. Check the last lines.

Thanx

Here is the complete code, it reads a table with utm coordinates, then project these coordinates along a profile obtaining a plot:

type of data in the table:

ANTI 1 1 640772 5786590 1057.66 37.08 1.31 classic 735560.77881
ANTI 1 2 641584 5786571 1870.49 133.37 2.65 classic 735560.77924
ANTI 2 1 641108 5785042 899.36 43.10 4.30 classic 735560.77969
CONI 3 1 734515 6087409 798.25 45.00 0.38 classic 735566.41527
CONI 2 1 737715 6087097 1359.38 50.00 0.81 classic 735566.41698
CONI 2 2 737374 6087613 1977.23 75.00 1.10 classic 735566.41894
CONI 4 2 733623 6086157 422.26 62.07 2.79 classic 735566.41973

and this is the script:

      %projection of shorelines routines
      %load file
      file=('shorelines.txt');
      if exist(file,'file')==1 || exist(file,'file')==2

      %load shorelines table
      fid=fopen(file,'r');
      i=1;
      while ~feof(fid)
          %tline = fgetl(fid);
          A=textscan( fid,'%s %u %u %f %f %f %f %f %s %f/' ...
                    , 'delimiter',' ','HeaderLines',1);

          for j=1:10 %number of columns (default)
              data(i,j) = A(1,j);
          end
          i=i+1;
      end
      fclose(fid);
      %extract data from table
      nim=numel(data(:,10));
      u=1;

      for k=1:nim
      md1(u,1)=data{k,4}; %x
      md1(u,2)=data{k,5}; %Y
      md1(u,3)=data{k,7}; %z
      md1(u,4)=data{k,8}; %ze
      md1(u,5)=data{k,3}; %shore
      u=1+u;
      end
      end

      % set variables 
      x=md1(:,1);
      y=md1(:,2);
      z=md1(:,3);
      ze=md1(:,4);
      shore=md1(:,5);

      figure(1) %input figure

      clf
      hold on
      scatter(x,y,75,z,'filled');
      col=colorbar;
      tite = get(col,'title');
      set(tite,'string','Elevation (m)');

      axis equal
      box on

      ylabel('UTM   N')
      xlabel('UTM   E')

      %graphic input
      [xl,yl]=ginput(2);
      l(:,1)=xl;
      l(:,2)=yl;

      plot(xl,yl,'-k') %plot of profile
      plot(l(1,1),l(1,2),'or'); %point1
      plot(l(2,1),l(2,2),'ob'); %point2

      dx=l(1,1)-l(2,1);
      dy=l(1,2)-l(2,2);
      m=(dy/dx);
      m2=-1/m;
      an=-atan(m)*180/pi;
      anr=an*pi/180;
      ddg=sqrt((l(1,1)-l(2,1))^2+(l(1,2)-l(2,2))^2);
      dkm=(ddg);
      b=l(1,2)-m*l(1,1);

      %text in plot
      str1=num2str(dkm/1000); %m to km
      text( min(xl)+(dx/2),min(yl)+(dy/2) ...
          , sprintf('%s km',str1),'FontSize',12)

      %rotate and reproject data

      for i=1:length(md1(:,1))
          b2=md1(i,2)-m2*md1(i,1);
          Pnx=(b2-b)/(m-m2);
          Pny=m*Pnx+b;
          Pn=[Pnx Pny];
          dp(i,1)=Pn(1,1); %xp
          dp(i,2)=Pn(1,2); %yp
          dp(i,3)=((l(1,2)-dp(i,2))/sin(anr)); % distance in km
          dp(i,4)=md1(i,3);
          dp(i,5)=md1(i,4);
      end

      % %%%%%%%figure B/N of shorelines
      figure(2)

      clf
      subplot(2,1,1)
      hold on
      box on
      %loop shorelines colors
      for i=min(shore):max(shore)      
         plot( dp(shore==i,3)/1000,dp(shore==i,4),'.' ...
             , 'color',[1/i 1/i 1/i],'markersize',20)
         j(i,1)=i;
      end
      %cartesian plane conditions for profiles
      if dx>0 && dy>0   
      xlim([-dkm/1000 0])
      a=1; %a is a parameter for data export
      elseif dx<0 && dy<0 
      xlim([0 dkm/1000])    
      a=3;
      elseif dx>0 && dy<0  
      xlim([-dkm/1000 0]) 
      a=4;
      else
      a=2;    
      xlim([0 dkm/1000]) 
      end
      %legend
      leg=legend(num2str(j),'location','eastoutside');
      tit = get(leg,'title');
      set(tit,'string','Shorelines');
      %plot errorbars
      errorbar(dp(:,3)/1000,dp(:,4),(dp(:,5)),'ok')
      ylabel('Elevation (m)')
      xlabel('Distance along profile (km)')

      % save routine

Answer 1

[David Young]

The conditions are almost mutually exclusive and almost correct. To be absolutely right, you need the second inequality in condition 2 and the first inequality in condition 4 to be > rather than >=. However this detail is unlikely to be causing your problem.

Nesting the if statements would not be helpful. Using elseif as suggested in Adam's comment would be good practice, as it makes the intention more explicit and is very slightly more efficient, but will make little difference to the behaviour (and will make no difference if you make the small corrections suggested above).

The problem is more likely to be in setting the value of the matrix l. You don't show the code that does that. I suggest that you look there, and also print out the value of l at the start of the conditional code, so that you can check what it is. I see that you plot values of dp - maybe l needs to be computed from dp somehow?

[Jules Ray]

thanx, instead of use the matrix l, i use the difference between their elements and work properly now.