Crossfilter and dc.js mean value

Question

I have this data set:

tick    ID    Name    val1    val2    val3    val4
0       1     Name1    0.7    0.1       0.2    0.1
1       2     Name2    0.8    0.1       0.1    0.1
2       3     Name3    0.4    0.1       0.2    0.3
3       4     Name4    0.6    0.1       0.2    0.1
4       1     Name1    0.7    0.1       0.2    0.1
5       2     Name2    0.8    0.1       0.1    0.1
6       3     Name3    0.4    0.1       0.2    0.3
7       4     Name4    0.6    0.1       0.2    0.1

I already visualized everything with crossfilter and dc.js, except one diagram. I want the mean value of the column val1, val2, val3 and val4.

I have visualized this map with dc.js alread. The ID in the data is a segement in the image. The ticks are rounds, for instance the object with id 1 at tick 2 has the values 0.7, 0.1, 0.2 and 0.1. So the mean data should be connected to the segments. For instance when only the blue object with ID 1 is selected it should show the mean of the current selection:

tick    ID    Name    val1    val2    val3    val4
0       1     Name1    0.7    0.1       0.2    0.1
4       1     Name1    0.7    0.1       0.2    0.1

If I select a second object with ID 2, it should show the mean value of the both segments:

tick    ID    Name    val1    val2    val3    val4
0       1     Name1    0.7    0.1       0.2    0.1
1       2     Name2    0.8    0.1       0.1    0.1
4       1     Name1    0.7    0.1       0.2    0.1
5       2     Name2    0.8    0.1       0.1    0.1

Maybe I can show then the mean values of the two object in an pie chart. Where I can see the mean of the four classes val1, val2, val3 and val4. So I want to set them in relation like:

Sum val1 3
Sum val2 0.4
Sum val3 0.6
Sum val4 0.4

Sum all 4.4

So the classes would have the following size on the pie chart:

val1 3/4.4 = 0.68 = 68 percent
val2 0.4/4.4 = 0.090 = 9 percent
val3 ...
val4 ...

How can I accomplish this, because this was the only visualization that didn't worked.

Answer 1

It's funny how many different ways people explain how their data is structured and the result they need. Even with great detail explanation sometimes the message doesn't come across. Anyway I heard this post as being similar to my post . Gordon suggested I use reductio . However I dug into some other posts I ended up using the add remove initial methods. I created a jsfiddle that gives you the percent per name. I would like to know from others more skilled with DC if this is being done correctly, or not.

   var data = d3.csv.parse( d3.select("pre#data").text() );     
            data.forEach(function (d) {
            d.tick = +d.tick
            d.id = +d.ID;
            d.name = d.Name;
            d.val1 = +d.val1;
            d.val2 = +d.val2;
            d.val3 = +d.val3;
            d.val4 = +d.val4;
        });