How can i remove irrelevant information in plotting

1 次查看（过去 30 天）

Abdul Sajeed Mohammed 2022-2-19

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1653690-how-can-i-remove-irrelevant-information-in-plotting

评论： Voss 2022-2-20

Hello, i have plotted for error but seems like i have been getting a lot of noise in image. i want to only get the "Orange" highlighted path instead of shaded "grey-black" part. Any suggestions enclosing image below.

9 个评论
显示 7更早的评论隐藏 7更早的评论

Abdul Sajeed Mohammed 2022-2-19

编辑：Walter Roberson 2022-2-19

在 MATLAB Online 中打开

Ok, i hope i can explain my problem here,

Estimated_Lat = 180 / pi * lat;           %Dimension of (1x17341)
Estimated_Long = 180 / pi * lon;       %Dimension of (1x17341)
figure;
plot(Estimated_Long, Estimated_Lat,'x','Linewidth',0.5);
hold on
Ground_Truth = xlsread('27');
G_Long = Ground_Truth(:,1);         (43x1)
G_Lat = Ground_Truth(:,2);            (43x1)
plot(G_Long,G_Lat,'-x','Linewidth',2)
hold off
ylabel('Latitude, [deg]','FontSize', 13);
xlabel('Longitude, [deg]','FontSize', 13);
title('Trajectory of vehicle with position close to Lane level','FontSize', 14)
legend('Estimated trajectory', 'Centre nodes of True trajectory','best')
% To find error between estimated and true trajectory, i used difference as shown below
Error_dataLat = G_Lat - 180 / pi * lat;          (43x17341)
Error_dataLong = G_Long - 180 / pi * lon;   (43x17341)
plot(Error_dataLong,Error_dataLat) 

[ This is the image i have shared below, where instead of only highleghted area more of noise is available ]

Voss 2022-2-19

在 MATLAB Online 中打开

OK, so the figure is showing just Error_dataLat vs Error_dataLong (because hold is turned off after plotting the previous lines), and each of these are matrices of size 43-by-17341. When you use plot() with matrices, you get one line for each column of the matrix, so that's 17341 lines in this case (each with 43 data points). The line colors cycle through the axes' ColorOrder, and the lines are all on top of one another, so that's why the plot looks like a shaded grey-black region like that.

Rather than plotting those matrices as 17341 lines, maybe it is better to visualize them as two surfaces, something like this (I don't have your data, so I used random numbers but you can just comment out my surf(randn(...)) lines and uncomment the preceding lines that use your variables Error_dataLat/Error_dataLong, and see how that looks):

figure()

subplot(2,1,1);

% surf(Error_dataLat,'EdgeColor','none');

surf(randn(43,17341),'EdgeColor','none');

view([0 90]);

colorbar();

title('Lat Error');

subplot(2,1,2);

% surf(Error_dataLong,'EdgeColor','none');

surf(randn(43,17341),'EdgeColor','none');

view([0 90]);

colorbar();

title('Long Error');

Voss 2022-2-20

在 MATLAB Online 中打开

Recall, the errors are defined like this:

Error_dataLat = G_Lat - 180 / pi * lat;
Error_dataLong = G_Long - 180 / pi * lon;

So, yellow color in the surfaces means relatively high positive error (meaning G_Lat > 180 / pi * lat, or, respectively, G_Long > 180 / pi * lon) and blue means high (in magnitude) negative error (so G_Lat < 180/pi*lat, resp., lon).

However, usually error is defined to be in terms of absolute value of the difference if we don't care whether it's positive or negative and we just care about the magnitude, e.g.:

Error_dataLat = abs(G_Lat - 180 / pi * lat);
Error_dataLong = abs(G_Long - 180 / pi * lon);

and that might be more instructive to look at. In that case yellow will be high error using the new definition (i.e., high in magnitude, positive or negative, using the old definition) and blue will be low error (that is, closer to 0, so a better match).

I don't know what the 2 dimensions of the lat/long matrices represent, so I can't comment on what the "spatial" variation across the surface in X or Y might mean.

Abdul Sajeed Mohammed 2022-2-20