Overlaying scatter plot on grouped bar graphs

Hi All,

I am trying to plot a grouped bar graph. Along with the bars, I also need to plot single data points on each of the bars. Additionally, I would like to connect the individual data points with each other with a line. Here is the data I would like to use. Currently, I have been able to plot the bars and the error bars. But I am unsure about plotting the single data points. Is there a way to do that?

x = 1:3;
x = categorical({'High PU-Low RU', 'Low PU-High RU', 'Low PU-Low RU'});
SEM = [0.092290761	0.08898688	0.086731951	0.0825611
0.07279865	0.077586879	0.075656197	0.076483
0.088012243	0.088217074	0.094249201	0.081997937]
y = [0.619105248	0.596942136	0.581815617	0.553836698
0.488348187	0.520468608	0.507517199	0.513063563
0.590404075	0.591778124	0.63224286	0.550058883]
all = [0.1	0.2	0.3
0.4	0.4	0.6
0.6	0.5	0.8
0.6	0.8	0.6
0.1	0.2	0.3
0.4	0.4	0.6
0.6	0.5	0.8
0.6	0.8	0.6
0.42	0.52	0.62
0.72	0.72	0.92
0.92	0.82	1.12
0.92	1.12	0.92
0.42	0.52	0.62
0.72	0.72	0.92
0.92	0.82	1.12
0.92	1.12	0.92
]

Answer 1

[Star Strider]

The ‘all’ matrix is (16x3), however there are 12 bars (3 groups of 4). Ths size of ‘all’ is not consistent with that.

How is the ‘all’ matrix to be plotted with respect to the bars? I have chosen one option here. Clarify if I guessed incorrectly.

x = 1:3;
xlbl = {'High PU-Low RU', 'Low PU-High RU', 'Low PU-Low RU'};
SEM = [0.092290761	0.08898688	0.086731951	0.0825611
0.07279865	0.077586879	0.075656197	0.076483
0.088012243	0.088217074	0.094249201	0.081997937];
y = [0.619105248	0.596942136	0.581815617	0.553836698
0.488348187	0.520468608	0.507517199	0.513063563
0.590404075	0.591778124	0.63224286	0.550058883];
all = [0.1	0.2	0.3
0.4	0.4	0.6
0.6	0.5	0.8
0.6	0.8	0.6
0.1	0.2	0.3
0.4	0.4	0.6
0.6	0.5	0.8
0.6	0.8	0.6
0.42	0.52	0.62
0.72	0.72	0.92
0.92	0.82	1.12
0.92	1.12	0.92
0.42	0.52	0.62
0.72	0.72	0.92
0.92	0.82	1.12
0.92	1.12	0.92];

figure
hBar = bar(x, y.');                                                 % Return ,bar  Handle
for k1 = 1:size(y,2)
    ctr(k1,:) = bsxfun(@plus, hBar(k1).XData, hBar(k1).XOffset');   % Note: ;XOffset; Is An Undocumented Feature, This Selects The ‘bar’ Centres
    ydt(k1,:) = hBar(k1).YData;                                     % Individual Bar Heights
end
hold on
errorbar(ctr, ydt, SEM.', '.g', 'MarkerSize',1)    
plot(x, all, '.-', 'LineWidth',1.5)
xticklabels(xlbl);

.

[Prashanti Ganesh]

Hi thanks for this answer. I understand the inconsistency in the data. May be this data makes more sense. But I want the all data to be overlaid on each of the bars and then the line to be connected between the dots. In your solution plot, the dots are between the bars. Is there a way to change that? Thank you.

all = [1     2     4
     3     6     2
     7     4     2
     4     2     1
     3     5     4
     4     6     3
     4     3     2
     5     6     3]
y = [1     2     4
     4     6     2
     5     7     8
     2     4    10]

[Star Strider]

My pleasure.

These data still do not match the bar plot.

I have no idea how to plot them since the sizes (number of rows in ‘all’ and the number of bars) do not match, nor do I know what result you want. The bar centres (with respect to to x-axis coordinates) are provided by the ‘ctr’ matrix, created in the loop. Once the rows in ‘all’ match the number of bars (12), you can use those data to plot the lines in a loop. Just now, I have no idea how to do that, because the numbers do not match. I also have no idea how to match the rows of ‘all’ and the bars themselves, even if the sizes matched, in order to plot the lines. Those details — what rows of ‘all’ go with what bars — have not been provided.

.

[Prashanti Ganesh]

The reason for the inconsistency in the dimensions of all is because there are many data points whose avg is represented by one bar. For instance, in the following figure each bar represents the avg of the single data points overlaid on the bars. In this case, y was a matrix of 1 by 4 while all was a matrix of 50 by 4. I just want the same thing, for each of the bars, in a grouped bar graph. Even I am unsure what the dimensions of all should be like.

Answer 2

[Tala]

If I understand your question, you want to plot a few more points. just use plot(x,y,'o')

i added one point above the first bar in yellow.

[Prashanti Ganesh]

Yes, I want to add a few more points. But I want to add single data points for each of these bars. How can I do that?

[Prashanti Ganesh]

Basically, I want to plot the 'all' data as scatter points on the bars.