How to make chart of overlapping bell curves?

[Entropy]

I have a spreadsheet tracking weight loss efforts of a group. I find it motivating to look at the data lots of different ways. One thing that occurred to me that might be interesting is creating a sort of bell curve for each of us showing the distributions of our weekly loss/gains centered around 0. So the line is higher at -1 the more times you lost 1lb that week and so on. Different people’s scales sometimes track half pound or 2-tenths as a precision.

How can I make this happen? I’ve got various types of line charts, but I’m not sure where to start here. I am effectively wanting the value in the cell to be the x-axis, the count or proportion of how often they occur out of 100% as the y-axis, and I don’t want to have to sort the raw data set which is by date.

[Hagar_Delest]

If I understand correctly, you want the statistical distribution of the weight gain/loss, with the gain/loss as X axis, centered around 0 and in Y the % do that it makes a bell curve per normal law, correct?

The trick is that you need to carefully evaluate the number of “bins” and the width so that the rendering shows that bell curve. Have a look at: Number of bins and width, together with the rest of the article of course.

Once you’ve defined those 2 parameters, you have to create a table in which you’ll populate each bin with its count. So, you’ve to define the lower boundary (LB) and test the raw data (no need to sort it) to find out how many entries fit in the range [LB; LB+width[, then continue by adding the bin’s width each time until you reach the upper boundary.

Then, just calculate the % represented by each bin. The plot will use as X axis the bins and the Y axis the %.

Have fun!