There is no way that this can be normalised,
as such. Whatever you do, the spike at zero will
remain a spike of some kind.
Whether there is a case for treating this
as a zero-inflated distribution must be unclear
without knowing what the data are, but
as the non-zeros are not integers, I doubt
that the most popular solutions apply either.
Nick
[email protected]
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]]On Behalf Of fyzzz
>
> I need suggestion for data transformation
> Here is the ditribution of a measurement:
>
> snk_r1 Frequency Percent
> ------------------------
> 0 1750 91.62
> 0 ~ 0.04 55 2.88
> 0.04 ~ 0.08 55 2.88
> 0.08 ~ 0.12 22 1.15
> > 0.12 28 1.47
>
>
> This data is seriously biased to zero. I wonder
> whether there is any good method to transform
> this measurement and make its distribution into
> normal distribution, and how to do in stata.
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
