COMPGROUPS.NET | Search | Post Question | Groups | Stream | About | Register

### Re: the way impute non normal data

• Email
• Follow

```oslo <hokut1@YAHOO.COM> wrote
>
>I have data from molecular study (SNPs data). data set has 35000 columns and I have some missing data to be imputed. However I figured out from SAS documantaton that PROC MI is OK for�  rilmultivariate data (For data sets with arbitrary missing patterns, a Markov Chain Monte Carlo (MCMC) method (Schafer 1997) that assumes multivariate normality is used to impute all missing values or just enough missing values to make the imputed data sets have monotone missing patterns)
>
>Still can I use PROC MI to impute for my data?
>

Do you have monotone missing data?

If so, you can use other methods, designed for non-normal data.  If not, then you have to compromise somehow

Peter

Peter L. Flom, PhD
Statistical Consultant
Website: http://www DOT statisticalanalysisconsulting DOT com/
Writing; http://www.associatedcontent.com/user/582880/peter_flom.html
```
 0

See related articles to this posting

0 Replies
221 Views

Similar Articles

12/17/2013 10:16:16 PM
page loaded in 55122 ms. (0)

Similar Artilces:

the way impute non normal data
Dear User;=0A=0AI have data from molecular study (SNPs data). data set has = 35000 columns and I have some missing data to be imputed. However I figured= out from SAS documantaton that PROC MI is OK for=A0 rilmultivariate data (= For data sets with arbitrary missing patterns, a Markov Chain Monte Carlo (= MCMC) method (Schafer 1997) that assumes multivariate normality is used to = impute all missing values or just enough missing values to make the imputed= data sets have monotone missing patterns)=0A=0AStill can I use PROC MI to = impute for my data?=0A=0AHappy new year,=0A=0AOslo=A0=...

Re: How to transform Non-NORMAL data to NORMAL data in SAS? #3
You may consider Tukey’s Ladder of Transformations: http://davidmlane.com/teaching/503_2005/Tukey_ladder_transformation.pdf But again: Not all non-normal data can be transformed to be normal. Oleg ...

Re: proc MIXED and non normal data #3
Payet St�phanie <nanie.payet@LAPOSTE.NET> wrote: >I have fitted a random effect model with proc mixed, but the dependant >variable isn't normaly distributed. Can I trust the results ? Are there >other solutions ? >For example, for the model fitted on the non normal variable, the >treatment fixed effect isn't significant. But, if I fitt the model on the >log-transformed variable (which can be considered as normal) the >treatment >has now an effect. Is it legitimate to use log-transformed data to solve >the problem of normality ? [1] No, you cannot trust...

Re: proc MIXED and non normal data #4 1548170
I want to explain a score, which represents cumulated medical costs per p= atient (right skewed), with the treatment recieved and the length of sta= y in intensive care units. I used proc mixed because I wanted to take into account the fact that pat= ients are clustered into intensive care units. thanks for your help > >>> Payet St=E9phanie 6/22/2005 10:18:55 AM >>> wrote > <<< > Hello all ! > > I have fitted a random effect model with proc mixed, but the dependant > variable isn't normaly distributed. Can I trust the results ? Are there= &...

Transforming non-normal data to normal
I currently have a set of data which is non-normal. I would like to transform them to normal so as to plot a Hotelling T^2 graph that will allow me to monitor when the process goes out of control. My problem now is, I cannot seem to find any pre-defined functions in Matlab that will help me do so. Is there one that I missed out? Or are there better transformation methods I should look into? I believe Johnson transform is better than Box Cox from the research I have done. I appreciate your help! On Sep 28, 10:41=A0am, "Jo " <sk8te...@hotmail.com> wrote: > I currently have a ...

Comparing Normal Vs Non-normally distributed data
Hi All, I am working on two samples, out of which one is normally distributed and the other is not. I need to compare these two samples. I am not sure of which test to use but had done my comparison using Wilcoxon signed rank test. This test is usually used for non-parametric datasets. Can I use this to compare a parametric vs non-parametric ? On Mar 27, 8:21=A0pm, m.priya...@gmail.com wrote: > Hi All, > > I am working on two samples, out of which one is normally distributed > and the other is not. I need to compare these two samples. I am not > sure of which test to use but h...

Non Normal Data
Hi, I am looking for references or sources that discuss how to simulate non-normal multivariate data with a desired covariance structure ? If any of you know, can you direct me to those sources ? Regards. Zhu a simple and easy to read practical reference book is 'SAS for Monte Carlo Studies: A Guide for Quantitative Researchers' by Xitao Fan, et al. If you don't have this book, google key words: intermediate correlation procedure and related papers by Vale and Maurelli. You can use Google books to view a preview of this book. Since the book is published by SA...

Is there documented way to find out in a non-global zone that you're a non-global zone
Hello, I am looking for a documented way from Sun to tell in a non-global zone that I am running in a non-global zone. Thomas Hallo, > I am looking for a documented way from Sun to tell in a non-global > zone that I am running in a non-global zone. zoneadm list -cv seems to do what I want. In case of a global zone it says id 0 and "global" otherwise it says that it is id 1 and its configured name. Thomas On Nov 6, 9:26 am, Thomas Glanzmann <sithg...@stud.uni-erlangen.de> wrote: > Hello, > I am looking for a documented way from Sun to tell in a...

Re: How to normalize the data layout? 132576
Mark, If the need is only one time, and not a lot of variables are involved, sure, that approach can be more intuitive. However, I prefer the transpose approach as it naturally extends a lot easier to all of those conditions. Art ------- On Thu, 21 Jan 2010 18:01:26 -0500, Mark Miller <mdhmiller@GMAIL.COM> wrote: >I personally think a simple one-pass data step is a very intuitive >SAS solution for "unfolding"" or re-normalizing data since code pattern >makes it trivial to unfold multiple data items at the same time. > > >** sample ...