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

### URGENT missing value in PROC MIXED

• Email
• Follow

Hi everybody,

I want to use proc mixed, with this model ,

Y= x1 +x2 +x1*x2
question :
How proc mixed handle the missing value for estimating the effect.of
X1 or X2 when :
- Y is missing
- X1 or X12 is missing.
- both Y and X1 or X2 are missing
thanks,

 0

See related articles to this posting

On May 18, 6:21=A0am, sousimou <elidriss...@gmail.com> wrote:
> Hi everybody,
>
> I want to use proc mixed, with this model ,
>
> Y=3D x1 +x2 +x1*x2
> question :
> How proc mixed handle the missing value for estimating the effect.of
> X1 or X2 when :
> - Y is missing
> - X1 or X12 is missing.
> - both Y and X1 or X2 are missing

PROC MIXED will not use observations where the response Y is missing;

PROC MIXED will not use observations where one or more predictor
variables have missing values

--
Paige Miller
paige\dot\miller \at\ kodak\dot\com

 0

1 Replies
493 Views

Similar Articles

12/12/2013 7:49:15 PM
[PageSpeed]

Similar Artilces:

Re: proc format value assignment oddity
Paul: An 'empty' character value ('') does not appear to have any meaning in = PROC FORMAT (MSW SAS, V9.13). This program demonstrates that: proc format; value $miss ' ' =3D 'missing' '00'x =3D 'null '=20 other =3D 'notMiss' ; run; quit; data test; length x$ 10; x=3D''; output; x=3D' '; output; x=3D'00'x; output; x=3D'z'; output; run; proc freq data=3Dtest; tables x/missing; format x $miss. ; run; quit; It does not seem to matter how many blanks one uses to make a missing = character... Missing data in Proc Mixed I am responding to a reviewer about the use of proc MIXED in a repeated measures ANOVA with treatment and control. We think that we are employing an Intent To Treat method because Proc mixed does not employ listwise deletion as does proc GLM. However the reviewer thinks that Proc Mixed must be substituting values. I notice that most authors simply say that because Proc Mixed using ML estimation, that it employs all of the data and cases are not eliminated if they provide at least one data point. My question is, "Can someone provide me with either an explanation for how the ML algorithm... Proc TABULATE, mixed up on missing I am using the MISSING option so that a TABULATE will occur without getting a WARNING: A class, frequency, or weight variable is missing on every observation. and no output at all. Is there a way to use the MISSING option and then tell tabulate to not show the missings (class level values)? In this sample, I would want to see just columns N X*1*N X*2*N Z*1*N data foo; do topic = 1 to 2; do id = 1 to 100; r = ranuni(1234); select ; when (r < 0.33) x = .; when (r < 0.66) x = 1; otherwise x=2; end; y = .; z = ifn (ranuni(1234)<0.2,.,1); ... Exclude missing values in proc freq Hi, Is it possible to tell proc freq to exclude missing values in the created dataset with an option. Or will i have to use a where clause in the output dataset data one; input a @@; cards; 1 1 1 1 1 . . .1 ; run; proc freq data=one noprint ; table a / out =two ; run; On Jun 16, 12:45=A0pm, inason <iharu...@gmail.com> wrote: > Hi, > Is it possible to tell proc freq to exclude missing values in the > created dataset with an option. > Or will i have to use a where clause in the output dataset > > data one; > input a @@; > cards; > 1 1 1 1 1 =A0. . .1 > ; &... HGLM: PROC MIXED + PROC LOGISTIC? Dear all, Does anybody know a command for Hierarchical General Linear Modeling (HGLM)? I am sure that a SSI's software, called HLM, can do hierarchical non-linear modeling, but I would like to work with the data in SAS if possible. I would appreciate if any of you could show me an example command when I have a dichotomous (or categorical) dependent variable (dv), a level-one independent variable (lv1), a level-two independent variable (lv2), and a id code over two levels (id) as below: PROC MIXED DATA=hlm.educ IC; CLASS id; MODEL dv= lv1 lv2 lv1*lv2 / SOLUTION; RANDOM intercept lv1 /... PROC MIXED OR PROC MODEL (GMM) Hi all, I would like to analyze the effects of within "subjects" and between "subjects" in a datasets with the random and/or fixed effects. And, I realize that the data (panel data) may contain the autocorrelation and heteroskedasticity problems. So, If we would to reduce the autocorr. and heteroskedasticity problems, should PROC MIXED or PROC MODEL be used? If GLS is normally used to reduce the heteroskedasticity problem, does it minimize the autocorr. problem as well? In this case, which SAS PROC statement is much more suitable? Thank you very much! Thomas ... Re: RCB with subsampling in Proc GLM and Proc Mixed add the following: random block block*trt; Mixed should be able to figure out you want BxT as the error term for trt. It's always a good idea to check if ddf corresponds to your expectations. On Fri, 7 Apr 2006 09:37:17 -0700, Nancy Koerth <nkoerth@YAHOO.COM> wrote: >I have used the following program using Proc GLM. However, I would like to take advantage of Proc Mixed's ability to handle heterogeneous variances. How do I compose the program in Proc Mixed? > > I have 4 blocks (fixed), 5 trts, and 5 subsamples/unit. > > proc glm ; > class block trt; >... missing values I want to read a csv file that contains numerical values and some values are missing. I can handle it by reading std::strings (separated by commas) and then converting them to doubles (if they are not empty). I was wondering if there is a better way. If I try to do this const double NAV = -999; double ReadDouble(ifstream &in) { double readValue = NAV; in >> readValue ; return readValue ; } it doesn't work. The very first time ifstream doesn't contain a number, then readValue is still set to NAV (which is what I want) but from that point ifstream is all messed up and I c... format row value based on value of another column proc report? Hi, Is it possible to format values in proc report based on the value in another column. For example, in a report for each company I have several measures, some are in percent and some are dollars. I would like to report any "cost" or "expense" with a dollar format and percents with percent format. As my code stands now, I can do only dollar format or a percent format. I guess I need to pre process but I'm not sure where to begin. Thanks for any insights. data test; input measure :$40. fY_2006; infile datalines dsd; datalines; % Contracted expenses, 1.4 % of Am...

Missing values 9378
Hi, In the code below i have a problem with line "if dat EQ &datum then " because this instruction doesn't work. I have missing values for v1 and v2. I don't see what's wrong here. Tx options COMPRESS = YES ; options OBS = MAX ; options MPRINT NOCENTER ; data _null_ ; %let datum2 = 01nov2009 ; %let datum = %sysfunc(inputN(&datum2,date9),yymmN6) ; run ; filename nnm 'c:\inp.txt' ; data data.inp ; infile inp obs = 1000 ; input @25 dat 6. @ ; if dat EQ &datum then input @1 v1 $8. @9 v2$16. ; run ; ...

A proc mixed question
Hi, I have a dataset consists of 550 rats who were randomly selected to two control and three dose groups. There were 55 males and females in every group. Each rat got a daily dose of the same concentration (air or low,medium ,high) throughout the study and was weighted weeky during a period of 61 weeks. That is, at the termination of the study each rat should have 61 measurements. However some rats donot have all 61 measurements, some have 42 measures, others have 54 and so on. One of my objectives is to investigate if there are differences between control and dose groups, therefore I chose t...

Fill missed values
I have a data set containing recorded values over fixed intervals. However, some intervals are skipped when no value is recorded. How can I add back those skipped intervals and fill them with the immediate previous recorded value? Below is an example of what I have and what I want. Currently the 3rd interval for day 2is skipped. I want to add the 3rd interval for day 2 and fill it with value from interval 2 on day 2, which is the immediate previous recorded value. Appreciate for your time and help. ------- data I have ---- data raw; input day intv value; cards; 1 1 0.1 1 2 0.3 ...

How to transfer all missing value all at once?
I have a data set with many missing value both in "." or " ", how can I transfer them all at once? I used the code below but return 0 observations in the new dataset. Thank you in advance for your help. Melody On Mar 15, 2:48=A0pm, Melodyp <pearsonmel...@gmail.com> wrote: > I have a data set with many missing value both in "." or " ", how can > I transfer them all at once? > > I used the code below but return 0 observations in the new dataset. > > Thank you in advance for your help. > > Melody Missed your "code bel...

interpolate with missing values
Hi IDLers, I would like to have a 2D interpolation routine to do the same as INTERPOLATE but taking in account missing values. What I am doing right now is setting the missing values to !NAN and then I use INTERPOLATE. When the routine wants to interpolate a point surrounded by 4 points, it uses bilinear interpolation and if one of them is !NAN then it returns !NAN. In this same example (1 point = !NAN) I would like to compute the interpolation using some kind of triangulation using the 3 finite and valid points. I don't really know if this way to proceed is correct and what I should do ...

missing value #3
Hi all; I have the following data; and I want to replace the missing value of the variable mtb by the first not missing one when the variable ri is not missing. date Id_firm ri mtb 01/01/1990 34567 . . 01/02/1990 34567 23 . 01/03/1990 34567 25 . 01/04/1990 34567 30 2 01/05/1990 34567 35 1,5 01/06/1990 34567 33 1,7 01/01/1990 14493 . . 01/02/1990 14493 60 . 01/03/1990 14493 41 1,5 01/04/1990 14493 32 2 01/05/1990 14493 56 1,9 01/06/1990 ...

PROC MIXED #5
Can someone tell me whether PROC MIXED estimates the fixed effect coefficients and random effect coefficients using the mixed model equations for repeated measures data and/or when the random effects are nested? I ask this question because the mixed model equations are not easy to specify in certain situations. For example, for multiple trait REML the MME are difficult to specify using multiple diagonalization for more than one random effect. Any help including references are much appreciated. Hars Hars, Your question is not clear. I believe that the answer to your question is yes,...

proc factor with missing data?
Does anyone know how to run an exploratory factor analysis presumably with PROC FACTOR that is able to deal with missing data? Is it possible to have output from PROC MI be read into PROC FACTOR? Or some other way of solving the problem? We are trying to do something similar as the following paper was able to do in SPSS. Personal author, compiler, or editor name(s); click on any author to run a new search on that name. <http://www.eric.ed.gov:80/ERICWebPortal/Home.portal?_nfpb=true&_pageLabel=E RICSearchResult&_urlType=action&newSearch=true&ERICExtSearch_Searc...

R square in proc mixed?
Hi all. How can I obtain R square in proc mixed or maybe something similar such that I can use in order to know how much variability is explained by the mixed model? Have a good day. Many thanks in advance. ...

Optional values in a Stored Proc
The following SP gives an error of: Server: Msg 245, Level 16, State 1, Procedure spSelectSEICData, Line 26 Syntax error converting the varchar value '@' to a column of data type int. In the Procedure I am using the Select * for testing purposes. Here is the proc. CREATE PROCEDURE spSelectSEICData ( @IndivNo int, @CommType SmallInt, @BeginDate as SmallDateTime ) AS Declare @SqlStr as char(1) Set @SqlStr = '' If ((@BeginDate <> ' ') and (@CommType <> ' ')) Begin Set @SqlStr = '@IndivNo AND [SEIC-COMMENT-TYPE] = @CommType' End I...

Re: Missing values #20
fkhurshed@HOTMAIL.COM replied: > >On Apr 19, 10:13 am, mkek...@YAHOO.CO.UK (Marina Kekrou) wrote: > > I have a sample including approximately 1,500 firm-year observations. > > Although the sample was much larger there were a lot of missing > > observations. Except from deleting missing observations, what other > > strategies have users found useful? > > > > I would appreciate any ideas or feedback on this. > >Some obvious ones, fill in with the average or median value. If it's >a time series, possibly interpolate for that point with regre...