Plotting and labelling without repetition a scatter /line plot in matlab.

hi!

I am trying to plot and label line/scatter plot of two columns in a dataset , the first column are the time values to be plotted on x - axis and the corresponding pitch values are in the second column of the dataset to be plotted on the y-axis. The labels for each point are in third column of the dataset labelled as notation. The labels should be added as such only the first value of repeated entries should be taken as label. The columns of dataset are 'char' array and the length of dataset is x 6839*3. An example of dataset is given below:

Time Pitch notation

2.64 329.63 'GM'

2.64 329.63 'GM'

3.49 261.63 'SM'

4.09 349.23 'MM'

6.27 261.63 'SM'

6.50 349.23 'MM'

6.72 329.63 'GM'

6.75 349.23 'MM'

7.02 329.63 'GM'

7.28 349.23 'MM'

7.32 329.63 'GM'

8.06 261.63 'SM'

8.44 329.63 'GM'

9.36 261.63 'SM'

9.95 349.23 'MM'

11.81 329.63 'GM'

12.04 246.94 'NL'

12.09 261.63 'SM'

12.29 329.63 'GM'

when i use the following commands , i do get a graph with labels but repeated entries are not removed as of now: ds = dataset('XLSFile', 'Deshgatannotation.xlsx','sheet', 2);

figure

time = ds(:,1);

pitch = ds(:,2);

l = cellstr(ds(:,3));

plot(time, pitch,'.','color','r')

text(double(time), double(pitch), l,'fontsize',6, 'color','b','rotation',45,'horizontal','left', 'vertical','bottom');

xlabel('time');

ylabel('notation');

title('notations vs time')

[dpb]

So by "repeated entries" do you mean only one label of each unique ID or only remove the case where there are two (or more) of the same ID in sequence and start again w/ the next different ID?

If the former, unique will be your friend; if the latter, as I see your sample subset there would only be one point removed? In that case, using something like

ix=[true ~strcmp(l(2:end),l(1:end-1))]; text(t(ix),p(ix),l(ix))

should do. Will have to have t,p as valid numeric values, of course, and assumes l is the label in column cellstring array.

[avantika]

Hi! Yes i want the latter thing to happen i.e. only remove the case where there are two (or more) of the same ID in sequence and start again w/ the next different ID.

Using the following command gives an error message:

 ids=[true ~strcmp(l(2:end),l(1:end-1))];
Error using horzcat

Dimensions of matrices being concatenated are not consistent

[dpb]

You have row or column vector? I presumed row above; if column either use

ids=[true; ~strcmp(l(2:end),l(1:end-1))];  % column vector result

or

ids=[true ~strcmp(l(2:end),l(1:end-1))'];  % row result of input column

Have to either add a row or a column depending on which direction the label list is oriented or reorient the list to make the concatenation commensurate in dimensions. Which way specifically is dependent on your input direction and desired output.