# Make a differentiated barchart using numbers

I have a question about Numbers.

I want to make a bar chart that shows the membership of my organisation.

I want to divide between unitary organisations and membership organisations, so i want there to be 2 bars.

But on the bar for membership organisations, I want it to be divided by a categorisation of "how many members". The 4 categories would be <30, 30-100, 100-500 and 500+.

so for example the numbers would be

• Unitary organisations 3
• Membership organisations 11 (<30=2, 30-100=4, 100-500=3, 500+=2)
-
Sounds like you need a kind of histogram, but set out as a stacked column graph? – sblair Jul 28 '11 at 14:59
ya that looks about right, but i am not sure how to organise the data in the spreadsheet so that it all comes out in the right places... any hints? thanks – Joe Jul 28 '11 at 15:49
Shall the y-Axis show the number of members in total, so for example in the group below 30, there is one of 17 and one of 22 members, so the sum is 39 members and reaches that point on the y-Axis, or like in the sblair-graphic, where he only counts organisations. – user unknown Jul 28 '11 at 22:40

To expand on my comment, you can create histogram source data, as described here, such that the values are in the spreadsheet in a form like:

``````bin         frequency (unitary)   frequency (membership)
<30         2                     2
30-100      4                     3
101-500     3                     4
>500        2                     1
``````

Then select the cells and choose "Stacked Column" from the menu:

You will probably need to swap the axes to get the result you want, by clicking on "Switch Row/Column", as shown in the screenshot below:

The "100% Stacked Column" graph can show the relative composition of each type, rather than the absolute values.

-
He only wants one bar to be divided, afaik. And I would expect the number of members in total on the y-Axis, but there is not enough data in the example, to do that. – user unknown Jul 28 '11 at 22:40
@user Well spotted. In that case, deleting all values for the "frequency (unitary)" column except for those in the "<30" row should give that result. But, as you say, this could be about visualising the total members in each bin, which is a bit different. – sblair Jul 29 '11 at 1:15