Can we do the approximately unbiased estimator for the probability density function of cont. var

[Season]

Thanks for the code! By the way, given the size of the simulated sample is 123- a number not that large- it can serve as a good example of one of my previous theoretical explanation to @TomHsiung. The bandwidth of 2.5 is a little bit small in that while its selection leads to a kernel PDF estimate that reveals the general trend of a normal one, the curve is troubled by zigzag lines, a sign of noise.

[yabwon]

For that 123 sample size, H for 1 to 12 looks like this:

Bart

[Season]

Thanks! What a wonderful presentation! I think you can expand the explanations of your code, make it a research paper and submit it to SAS conferences or statistical journals.