crosshairs or just vertical line across linked axis plots

What I have are plots that are linked by the linkaxes() function. I would like to have a vertical line that spans all of the plots so that I can compare the peaks and valleys interactively between plots.

Answer 1

[Jiro Doke]

This may get you started:

function vertical_cursors

set(gcf, ...
   'WindowButtonDownFcn', @clickFcn, ...
   'WindowButtonUpFcn', @unclickFcn);

% Set up cursor text
allLines = findobj(gcf, 'type', 'line');
hText = nan(1, length(allLines));
for id = 1:length(allLines)
   hText(id) = text(NaN, NaN, '', ...
      'Parent', get(allLines(id), 'Parent'), ...
      'BackgroundColor', 'yellow', ...
      'Color', get(allLines(id), 'Color'));
end

% Set up cursor lines
allAxes = findobj(gcf, 'Type', 'axes');
hCur = nan(1, length(allAxes));
for id = 1:length(allAxes)
   hCur(id) = line([NaN NaN], ylim(allAxes(id)), ...
      'Color', 'black', 'Parent', allAxes(id));
end

     function clickFcn(varargin)
        % Initiate cursor if clicked anywhere but the figure
        if strcmpi(get(gco, 'type'), 'figure')
           set(hCur, 'XData', [NaN NaN]);                % <-- EDIT
           set(hText, 'Position', [NaN NaN]);            % <-- EDIT
        else
           set(gcf, 'WindowButtonMotionFcn', @dragFcn)
           dragFcn()
        end
     end
     function dragFcn(varargin)
        % Get mouse location
        pt = get(gca, 'CurrentPoint');
        % Update cursor line position
        set(hCur, 'XData', [pt(1), pt(1)]);
        % Update cursor text
        for idx = 1:length(allLines)
           xdata = get(allLines(idx), 'XData');
           ydata = get(allLines(idx), 'YData');
           if pt(1) >= xdata(1) && pt(1) <= xdata(end)
              y = interp1(xdata, ydata, pt(1));
              set(hText(idx), 'Position', [pt(1), y], ...
                 'String', sprintf('(%0.2f, %0.2f)', pt(1), y));
           else
              set(hText(idx), 'Position', [NaN NaN]);
           end
        end
     end
     function unclickFcn(varargin)
        set(gcf, 'WindowButtonMotionFcn', '');
     end
end

Save the above function. Here's an example of how to use it:

subplot(221);
plot(rand(10,1));
subplot(222);
plot(rand(10,2));
subplot(223);
plot([rand(10,1), rand(10,1)+1]);
subplot(224);
plot(rand(10,1));

vertical_cursors;

EDIT: As a response to a request, I added the ability to make the cursor disappear when clicking outside the axes. See the two lines of code with the comment "<-- EDIT".

[Paulo Silva]

1 vote, interesting function

[Mark Dragovan]

Exactly what I wanted. Thanks!

[Walter Roberson]

I notice this question getting asked more often these days. Perhaps we should create a FAQ question about this.

[Gina]

Very nice!

How would you program it so that when they click off the axes the cursor lines disappear?

[Jiro Doke]

@Gina, I modified my answer above to include that capability. I simply added two lines of code to make the line "disappear" upon clicking outside the axes.

[Gernot Druml]

Many thanks, this function makes my life easier !!!

Very good alternative to the removed 'fullcrosshair' since 2014b.

With the additional function 'linkprop' the cursor can be synchronised with additional figures.

[Robbin van Hoek]

@Jiro, many thanks for your answer perfect! I do however get an error when dragging the cursor:

Error using interp1 (line 257) The values of X should be distinct.

Error in vertical_cursors/dragFcn (line 48) y = interp1(xdata, ydata, pt(1));

Error while evaluating figure WindowButtonMotionFcn

Also is it also possible to link a crosshair in a scatterplot (based on position in array) So my data is displayed both in multiple subplots, which are now linked by a vertical line. but could you also link a crosshair in a scatterplot?

[charline tessereau]

@Giro thank you for this function.

I need to acces to the values of the cursor, how is it posible to obtain the values of different cursors (i need to calculate the time between 2 cursors) in an array for instance?

And how could i put the text under the x-axis? because the text is located at the right of the cursor which does not allow to see well the plot..

Sorry for my bad english, i hope I'll be understood, i can try an other way to express what i want to do if not.

[Ferd]

Hi,

I tried the example given and it works without issues. Then on another figure with 4 subplots that I've linked their x-axes, I've come across this error below,

Is there a way to get around this error message?

Error using griddedInterpolant

The grid vectors must contain unique points.

Error in interp1 (line 149)

F = griddedInterpolant(X,V,method);

Error in vertical_cursors/dragFcn (line 41)

y = interp1(xdata, ydata, pt(1));

Error in vertical_cursors/clickFcn (line 28)

dragFcn()

Error while evaluating Figure WindowButtonDownFcn.

Error using griddedInterpolant

The grid vectors must contain unique points.

Error in interp1 (line 149)

F = griddedInterpolant(X,V,method);

Error in vertical_cursors/dragFcn (line 41)

y = interp1(xdata, ydata, pt(1));

Error in vertical_cursors/clickFcn (line 28)

dragFcn()

Error while evaluating Figure WindowButtonDownFcn.

Error using griddedInterpolant

The grid vectors must contain unique points.

Error in interp1 (line 149)

F = griddedInterpolant(X,V,method);

Error in vertical_cursors/dragFcn (line 41)

y = interp1(xdata, ydata, pt(1));

Error while evaluating Figure WindowButtonMotionFcn.

Error using griddedInterpolant

The grid vectors must contain unique points.

Error in interp1 (line 149)

F = griddedInterpolant(X,V,method);

Error in vertical_cursors/dragFcn (line 41)

y = interp1(xdata, ydata, pt(1));

Error while evaluating Figure WindowButtonMotionFcn.

Error using griddedInterpolant

The grid vectors must contain unique points.

Error in interp1 (line 149)

F = griddedInterpolant(X,V,method);

Error in vertical_cursors/dragFcn (line 41)

y = interp1(xdata, ydata, pt(1));

Error while evaluating Figure WindowButtonMotionFcn.

Error using griddedInterpolant

The grid vectors must contain unique points.

Error in interp1 (line 149)

F = griddedInterpolant(X,V,method);

Error in vertical_cursors/dragFcn (line 41)

y = interp1(xdata, ydata, pt(1));

Error while evaluating Figure WindowButtonMotionFcn.

Error using griddedInterpolant

The grid vectors must contain unique points.

Error in interp1 (line 149)

F = griddedInterpolant(X,V,method);

Error in vertical_cursors/dragFcn (line 41)

y = interp1(xdata, ydata, pt(1));

Error while evaluating Figure WindowButtonMotionFcn.

Error using griddedInterpolant

The grid vectors must contain unique points.

Error in interp1 (line 149)

F = griddedInterpolant(X,V,method);

Error in vertical_cursors/dragFcn (line 41)

y = interp1(xdata, ydata, pt(1));

Error while evaluating Figure WindowButtonMotionFcn.

Error using griddedInterpolant

The grid vectors must contain unique points.

Error in interp1 (line 149)

F = griddedInterpolant(X,V,method);

Error in vertical_cursors/dragFcn (line 41)

y = interp1(xdata, ydata, pt(1));

Error while evaluating Figure WindowButtonMotionFcn.

[Matthias Luh]

Thank you very much for the code!

I extended it for datetime x-axis compatibility:

function vertical_cursors
    set(gcf, ...
       'WindowButtonDownFcn', @clickFcn, ...
       'WindowButtonUpFcn', @unclickFcn);

    % Set up cursor text
    allLines = findobj(gcf, 'type', 'line');
    hText = nan(1, length(allLines));
    for id = 1:length(allLines)
        hText(id) = text(NaN, NaN, '', ...
            'Parent', get(allLines(id), 'Parent'), ...
            'BackgroundColor', 'yellow', ...
            'Color', get(allLines(id), 'Color'));
    end

    % Set up cursor lines
    allAxes = findobj(gcf, 'Type', 'axes');
    hCur = nan(1, length(allAxes));
    for id = 1:length(allAxes)
        if isa(xlim(allAxes(id)), 'datetime') == 1
            x_lims = xlim(allAxes(id));
            %default_time = x_lims(1) + diff(x_lims) / 2; % Option A) use mid
            default_time = NaT('TimeZone',x_lims(1).TimeZone); % Option B) don't display at start
            nan_data = [default_time default_time];
        else
            nan_data = [NaN NaN];
        end
        hCur(id) = line(nan_data, ylim(allAxes(id)), ...
            'Color', 'black', 'Parent', allAxes(id));
    end

    function clickFcn(varargin)
        % Initiate cursor if clicked anywhere but the figure
        if strcmpi(get(gco, 'type'), 'figure')
            if isa(xlim(gca), 'datetime') == 1
                x_lims = xlim(gca);
                nan_time = NaT('TimeZone',x_lims(1).TimeZone);
                nan_data = [nan_time nan_time];
            else
                nan_data = [NaN NaN];
            end
            set(hCur, 'XData', nan_data);
            set(hText, 'Position', [NaN NaN]);
        else
            set(gcf, 'WindowButtonMotionFcn', @dragFcn)
            dragFcn()
        end
    end
    function dragFcn(varargin)
        % Get mouse location
        pt = get(gca, 'CurrentPoint');
        % Update cursor line position
        set(hCur, 'XData', [pt(1), pt(1)]);
        % Update cursor text
        for idx = 1:length(allLines)
            xdata = get(allLines(idx), 'XData');
            ydata = get(allLines(idx), 'YData');
            if isa(xlim(gca), 'datetime') == 1
                if pt(1) >= get_date_xpos(xdata(1)) && pt(1) <= get_date_xpos(xdata(end))
                    x_lims = xlim(gca);
                    x = pt(1) + x_lims(1);
                    data_index = dsearchn(get_date_xpos(xdata'),get_date_xpos(x));
                    x_nearest = xdata(data_index);
                    y_nearest = ydata(data_index);
                    set(hText(idx), 'Position', [pt(1), y_nearest], ...
                        'String', sprintf('(%s, %0.2f)', datestr(x_nearest), y_nearest));
                    % y = interp1(get_date_xpos(xdata), ydata, pt(1));
                    % set(hText(idx), 'Position', [pt(1), y], ...
                    %     'String', sprintf('(%s, %0.2f)', datestr(pt(1)), y));
                else
                    set(hText(idx), 'Position', [NaN NaN]);
                end
            else
                if pt(1) >= xdata(1) && pt(1) <= xdata(end)
                    y = interp1(xdata, ydata, pt(1));
                    set(hText(idx), 'Position', [pt(1), y], ...
                        'String', sprintf('(%0.2f, %0.2f)', pt(1), y));
                else
                    set(hText(idx), 'Position', [NaN NaN]);
                end
            end
        end
    end
    function unclickFcn(varargin)
        set(gcf, 'WindowButtonMotionFcn', '');
    end

    function xpos = get_date_xpos(x_value)
        ax1 = gca;
        % dx_days = diff(ax1.XLim)/24;
        x_min = ax1.XLim(1);
        xpos = datenum(x_value - x_min);
    end
end

Answer 2

[Jiro Doke]

Does the answers from this question help?

[Mark Dragovan]

It's not exactly what I had in mind, but is close. I would be happy with a vertical line that I can control with the mouse position that spans multiple plots.

Thanks for the reference to that question, it will get me started.

[arysteasz]

How to get matrix index from this code?

Answer 3

[Patrick Kalita]

How about just using the Pointer property on the figure?

f = figure;
subplot(2,1,1)
plot(rand(5))
subplot(2,1,2);
plot(magic(5))
set(f, 'Pointer', 'fullcrosshair')

[Walter Roberson]

That doesn't span multiple plots.

[Patrick Kalita]

I'm not sure what you mean. As long as the axes are arranged vertically in a figure (which isn't too unusual) it seems to do what was asked.

[Walter Roberson]

I see what you mean, Patrick. On the other hand, it would cover the entire figure rather than just the plots.

[Jiro Doke]

It seems that this solution is the easiest and cleanest when the axes are stacked vertically. I'm not entirely sure what the OP meant when he said "compare the peaks and valleys interactively between plots". Is it just visually line things up, or does he want more information like actual numbers.

Answer 4

[claudio]

Is it possible to display the values in a window (like in datacursormode, "window inside figure" display style)??