News for hackers. Stuff that matters.

A comment on Why you should be wary of relying on a single histogram of a data set. Parent comment

Re:Why you should be wary of relying on a single histogram o... by jfim on Friday April 12, 2013 @02:12AM (#5536925)
Indeed, but that estimate is likely to be less misleading in most cases than a histogram(which is just a uniform kernel that is always aligned with bin boundaries).
[ Parent | on HN ]
Re:Why you should be wary of relying on a single histogram o... by xfs on Friday April 12, 2013 @03:36AM (#5537135)
One particular parameter of the kernel, bandwidth, can result in highly misleading visualization given arbitrarily chosen values. Here is an example http://en.wikipedia.org/wiki/File:Comparison_of_1D_bandwidth...

The smoothing give unsavvy readers a false sense of accuracy. With histogram they can at least tell it's an approximation.

[ Parent | on HN ]
Re:Why you should be wary of relying on a single histogram o... by noelwelsh on Friday April 12, 2013 @06:34AM (#5537662)
Yup. Luckily there are good methods for choosing the bandwidth:

http://www.umiacs.umd.edu/labs/cvl/pirl/vikas/Software/optim...

[ Parent | on HN ]