- #1
Saladsamurai
- 3,009
- 7
Hey people,
I am trying to write my own histogram making code since the Excel one sucks IMO. I am having some trouble thinking the whole bin 'width' thing through though.
I plan on using the formula
Then I was going to figure out the bin width as
This will get me in the ball park, but it will result in bin widths that are sloppy like 1.111 - 2.222 and the like. They should be sensible intervals like: 0 - 10, 10 - 20... etc or 0.10 - 0.12, 0.12 - 0.14...etc
Anyone care to brainstorm on me with this? I was thinking I could do some sort of test on the data set to determine what decimal place the bin width needs to go out to.
Hopefully someone can decipher what I am talking about here and wants to jump on board!
I am trying to write my own histogram making code since the Excel one sucks IMO. I am having some trouble thinking the whole bin 'width' thing through though.
I plan on using the formula
number of bins = k = 1.87(N - 1)0.40 + 1
Then I was going to figure out the bin width as
\frac{[MAX(data ) - MIN(data )]}{k}
This will get me in the ball park, but it will result in bin widths that are sloppy like 1.111 - 2.222 and the like. They should be sensible intervals like: 0 - 10, 10 - 20... etc or 0.10 - 0.12, 0.12 - 0.14...etc
Anyone care to brainstorm on me with this? I was thinking I could do some sort of test on the data set to determine what decimal place the bin width needs to go out to.
Hopefully someone can decipher what I am talking about here and wants to jump on board!
Last edited: