MATLAB Answers

Scatter not ignoring NaN in image

57 views (last 30 days)
H ZETT M
H ZETT M on 1 Aug 2017
Commented: Guillaume on 23 Sep 2017
Hey everyone, I need your help for a short moment. I used "scatter" to plot the following image.
I am quite happy with the image, but there is one problem. The data that I circled are NaN, but they simply get displayed as the lowest value by the scatter function. I tried to change the colorbar into ignoring NaN or making them invisible, but this did not work. Do you know a way to make the NaN values invisible (or simply remove them ?). The matrices I used to create this image are 11x11.

  3 Comments

Guillaume
Guillaume on 1 Aug 2017
It would be useful to know which of the inputs of scatter are NaN (The 4th input?) and exactly how you are invoking scatter.
The following works fine for me:
[x, y] = meshgrid(1:10);
c = 1:100;
c([10:10:40, 19, 29]) = nan;
scatter(x(:), y(:), [], c(:), 'filled');
The NaNs are not plotted (R2017a)
H ZETT M
H ZETT M on 1 Aug 2017
Hey, thank you for trying to help me. So I wrote some wrong stuff in the question (super tired). So my "code" is
scatter(alpha,beta,500,distance,'filled')
alpha and beta are both vectors with 121 entries (is this the right word ?). distance is also a vector with 121 entries and contains the NaNs. I used 500 for the size. I am currently working with matlab 2015b.
H ZETT M
H ZETT M on 1 Aug 2017
Oh well, I just figured out, that simply removing the NaNs while removing the entries from alpha and beta works fine. Thank you for your help anyway.

Sign in to comment.

Accepted Answer

Guillaume
Guillaume on 1 Aug 2017
As mentioned, the NaNs are correctly ignored in R2017a. I don't have R2015b installed anymore to test. You can easily work around the issue by removing the NaNs yourself:
nonans = ~isnan(distance);
scatter(alpha(nonans), beta(nonans), 500, distance(nonans), 'filled')

  0 Comments

Sign in to comment.

More Answers (1)

Muhammad RSMY
Muhammad RSMY on 23 Sep 2017
Data(isnan(Data))= 0;
idx = find(Data);
[X, Y] = ind2sub(size(Data), idx);
pointsize = 40;
scatter(X(:), Y(:), , pointsize, Data(idx),'square','filled');
colormap jet
colorbar

  1 Comment

Guillaume
Guillaume on 23 Sep 2017
Not sure why you're answering this 2 months after the question has been asked, particularly without any explanation.
Data(isnan(Data))= 0;
idx = find(Data);
[X, Y] = ind2sub(size(Data), idx);
Well, that's a very convoluted way of simply doing
[row, col] = find(~isnan(Data));
Note that conventionally row = Y and col = X.

Sign in to comment.

Products

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!