How can I classify data in and print it, all in real time?

My code is based on an answer I saw here.

I'm using mobile matlab app and I'm trying to classify the Azimuth sensor to 3 groups: left, middle and right- based on the Azimuth value. However, when I'm trying to print it after classification using drawnow, the plot ignores it and shows the actual value, not the classification.

 %connect to phone and get accel data
clear
clc
clear m
m = mobiledev;
m.AngularVelocitySensorEnabled = 1;
m.OrientationSensorEnabled = 1;
m.Logging = 1;
%initialize data for rolling plot
zvel = zeros(2000,1);
Azimuth= zeros(2000,1);
%initialize plot
figure(1)
hold on
p = plot(zvel);
p1=plot(Azimuth);
axis([0 200 -30 30]);
pause(1)
i=1;
while (m.Logging ==1 ) %runs until manual stop
      [av,~] = angvellog(m);
      [o,~] = orientlog(m);
      if length(av) > 200
        zvel = av(end-199:end,1);
      else
        zvel(1:length(av)) = av(:,1);
      end
      if length(o) > 200
        Azimuth = o(end-199:end,1)+135;
 %added 135 because idle azimuth=-135, and I wanted it to be around zero.
      else
        Azimuth(1:length(o)) = o(:,1)+135;
     end
      % redraw plot

      if (Azimuth(i)<=20 && Azimuth(i)>-20)
          Azimuth(i)= 0; %middle
      elseif (Azimuth(i)<=90 && Azimuth(i)>20)
          Azimuth(i) = 10; %right
      elseif (Azimuth(i)<=-20 && Azimuth(i)>-90)
          Azimuth(i) = -10; %left
      end
      p.YData = zvel;
      p1.YData=Azimuth;
      drawnow
      i=i+1;
end
hold off

I'm not an expert so I guess things here could be a lot better. However, solving the problem I mentioned is my main concern.

Thanks,

Netanel