Sturges' Formula

How can I prove Sturges' formula for determining number of classes in frequency distribution? k=Log2n+1 (n=number of data, k=number of classes)


Ambassador to the humans
You'll need to be more specific. Sturges' Formula isn't a theorem - it's a rule for how many bins to use. It may or may not give the "optimal" number of bins (for some measure of optimal) depending on the actual underlying data. But we aren't mind readers so you'll have to be more specific.


Omega Contributor
Oooo, that is an old paper. Those are fun sometimes since they seem to be straight to the point. Yes, BGM has a good point on finding the source document. I am sure there are also numerous comparison papers between methods that illustrate the pros and cons.