ryah (![[info]](http://l-stat.livejournal.com/img/userinfo.gif){kind=small} four) wrote,@ 2008-05-08 15:29:00 |
|
question
What is a good way to visualize a one dimensional data set? Perhaps for a list of dates in time line like a list of commit times into an SCM. The obvious way is to plot dots on a line but often I think it isn't so easy to see the distribution.
Perhaps plotting a density function?
E.g. You could pick a resolution window and count the points in that. The left hand end point of that window would be the input of the function -- the count would be the output. Maybe something a bit more complicated with a normally weighted average?
Any other suggestions? Surface plot of density function crossed with the window resolution?
What is a good way to visualize a one dimensional data set? Perhaps for a list of dates in time line like a list of commit times into an SCM. The obvious way is to plot dots on a line but often I think it isn't so easy to see the distribution.
Perhaps plotting a density function?
E.g. You could pick a resolution window and count the points in that. The left hand end point of that window would be the input of the function -- the count would be the output. Maybe something a bit more complicated with a normally weighted average?
Any other suggestions? Surface plot of density function crossed with the window resolution?
