The following code produces the image shown:
probabilities = datasetlist(1,:);
avgscores = datasetlist(2,:);
x = probabilities;
y = probabilities;
err = avgscores;
hold on
for k = 1:length(x)
e1 = errorbar(x(k),y(k),err(k),'-');
if err(k) == min(err)
set(e1,'Color','r')
set(e1,'MarkerEdgeColor','r')
set(e1,'Marker','*')
else
set(e1,'Color','k')
set(e1,'MarkerEdgeColor','k')
set(e1,'Marker','.')
end
end
hold on
e1.LineStyle = '-';
But, there should be a line connecting the datapoints. I even set the e1.LineStyle
, but that didn't work. How can I produce that line?
You have no line because you don't plot vectors, but single values each time, which is why you get something that is closer to a scatter
plot.
Below are two of the ways to fix this:
Solution 1
is a workaround that changes your existing code minimally.Solution 2
is much shorter code that achieves a same-looking result, by plotting vectors directly. (recommended).
function q40765062
%% Prepare the data:
datasetlist = [0.4:0.05:0.7; abs(randn(1,7))];
probabilities = datasetlist(1,:);
avgscores = datasetlist(2,:);
x = probabilities;
y = probabilities;
err = avgscores;
%% Solution 1:
figure();
hold on
for k = 1:length(x)
e1 = errorbar(x(k),y(k),err(k),'-');
if err(k) == min(err)
set(e1,'Color','r')
set(e1,'MarkerEdgeColor','r')
set(e1,'Marker','*')
else
set(e1,'Color','k')
set(e1,'MarkerEdgeColor','k')
set(e1,'Marker','.')
end
end
plot(x,y,'-k'); % < The main difference in this solution.
%% Solution 2:
figure();
[~,I] = min(err); % < We compute the index of the minimal value before plotting anything.
errorbar(x,y,err,'-*k'); hold on; % < Notice how we use the entire vectors at the same time.
errorbar(x(I),y(I),err(I),'-*r'); % < We only plot one value this time; in red.