Monday
StatDave
SAS Super FREQ
Member since
06-23-2011
- 1,724 Posts
- 2 Likes Given
- 390 Solutions
- 1,809 Likes Received
-
Latest posts by StatDave
Subject Views Posted 718 Monday 852 Sunday 797 Saturday 443 2 weeks ago 212 2 weeks ago 267 2 weeks ago 617 3 weeks ago 148 3 weeks ago 714 3 weeks ago 362 02-17-2025 12:26 PM -
Activity Feed for StatDave
- Got a Like for Re: Appropriate model for non-normal distribution. Monday
- Posted Re: Appropriate model for non-normal distribution on Statistical Procedures. Monday
- Got a Like for Re: Appropriate model for non-normal distribution. Monday
- Posted Re: Appropriate model for non-normal distribution on Statistical Procedures. Sunday
- Posted Re: What test should I use? on Statistical Procedures. Saturday
- Got a Like for Re: Determining which to use: Exp(est) or Odds ratio. a week ago
- Got a Like for Re: Joint mean and variance modeling in SAS: any suggestions on which PROCs should I use?. a week ago
- Got a Like for Re: Count data over years and repeated measures. a week ago
- Got a Like for Re: Fisher's - Taking long to run - 4x4 table with greater than 1000 sample size. a week ago
- Got a Like for Re: Joint mean and variance modeling in SAS: any suggestions on which PROCs should I use?. 2 weeks ago
- Posted Re: Joint mean and variance modeling in SAS: any suggestions on which PROCs should I use? on Statistical Procedures. 2 weeks ago
- Got a Like for Re: Count data over years and repeated measures. 2 weeks ago
- Posted Re: Count data over years and repeated measures on Statistical Procedures. 2 weeks ago
- Got a Like for Re: Fisher's - Taking long to run - 4x4 table with greater than 1000 sample size. 2 weeks ago
- Got a Like for Re: Fisher's - Taking long to run - 4x4 table with greater than 1000 sample size. 2 weeks ago
- Posted Re: Fisher's - Taking long to run - 4x4 table with greater than 1000 sample size on Statistical Procedures. 2 weeks ago
- Got a Like for Re: SAS two-part model. 2 weeks ago
- Got a Like for Re: SAS two-part model. 2 weeks ago
- Got a Like for Re: Extreme value of OR and HR and CI with FIRTH option in proc logistic and proc phreg. 3 weeks ago
- Posted Re: Extreme value of OR and HR and CI with FIRTH option in proc logistic and proc phreg on Statistical Procedures. 3 weeks ago
-
Posts I Liked
Subject Likes Author Latest Post 2 -
My Liked Posts
Subject Likes Posted 1 Monday 1 Sunday 2 2 weeks ago 2 2 weeks ago 3 2 weeks ago
06-02-2014
11:28 AM
Note that using DIST=NORMAL and LINK=LOG is not treating the response as lognormal, if that was your intention. To do that, model the log-transformed variable in PROC REG or PROC GLM. With these options in GENMOD, you are assuming that the response is normal and are modeling the log of the mean. If your response is positive, then your gamma model may be the more appropriate model.
... View more
05-16-2014
04:47 PM
Specify the RSQUARE option in the MODEL statement of PROC LOGISTIC to get both.
... View more
05-16-2014
10:46 AM
Also see this useful page which mentions neural networks: http://support.sas.com/kb/30333
... View more
05-16-2014
10:40 AM
3 Likes
Choosing a cutpoint depends on what criterion you want to use. The best tool for this is the CTABLE option in the MODEL statement. This option displays a table with statistics for each of a range of cutpoints such as the correct classification rate, false positive and negative rates, etc. If you want, you can change the cutpoints used in the table with the PPROB= option. For example, with this table you could find the cutpoint that maximizes the correct classification rate, or the cutpoint that satisfies your criteria for false positive and false negative rates.
There are other, follow on questions in this post. This response answers the original poster's question.
... View more
05-08-2014
05:05 PM
1 Like
Inverse confidence limits in the context of a binary response model (logistic or probit) can be done using the INVERSECL option in PROC PROBIT. The method used is described in the "Details: Inverse Confidence Limits" section of the PROBIT documentation.
... View more
05-08-2014
05:01 PM
I think you want b2+b3. The wage for single females under your model is b0+b1. For married females it is b0+b1+b2+b3. So the difference in wage is b2+b3.
... View more
04-02-2014
11:22 AM
A model on Y in GENMOD with LINK=LOG is not the same thing as a model on ln(Y) in REG. The former assumes Y is distributed as normal and models g(E(Y))=ln(E(Y)). The latter assumes ln(Y) is distributed as normal (Y is lognormal) and models E(ln(Y)). Generally, ln(E(Y)) is not the same as E(ln(Y)).
... View more
03-10-2014
05:01 PM
4 Likes
Rather than use a log-linked binomial model which can have the estimating problems that Jacob mentioned, you could try using a Poisson model with offset. See discussion of Zou's modified Poisson approach (and discussion of the log-linked binomial approach) in this note: 23003 - Estimating a relative risk (prevalence ratio)
... View more
02-28-2014
10:58 AM
1 Like
It sounds to me like this is a simple logistic model applied to aggregated data. So, you could use PROC LOGISTIC and events/trials syntax as below. Unless the data set is small and sparse, the EXACT statement isn't necessary and would probably take too much time. proc logistic; class time; model x/n = time; run;
... View more
02-27-2014
03:51 PM
It may be simpler to take a nonmodeling approach if you just have one predictor (type of fruit). This can be done with the CMH option in PROC FREQ. For example: proc freq; table strata*FruitType*response / cmh noprint; run; Of the three CMH statistics that are produced, the second one (Row mean scores differ) treats the row variable (FruitType) as nominal and the response as ordinal.
... View more
09-03-2013
01:48 PM
Model effect selection for generalized linear models is available beginning in the current release - SAS 9.4 TS1M0 - using PROC HPGENSELECT. For count models, it can also be done using the SELECTVAR= option in PROC COUNTREG in SAS/ETS software. See the documentation: http://support.sas.com/documentation/cdl/en/stathpug/66410/HTML/default/stathpug_hpgenselect_toc.htm http://support.sas.com/documentation/cdl/en/etsug/66100/HTML/default/etsug_countreg_toc.htm
... View more
01-31-2013
04:49 PM
1 Like
This note deals generally with using different formats in results -- it uses the example of changing precision, but the general idea of saving the data with ODS OUTPUT and printing with the desired formats applies generally. http://support.sas.com/kb/37106 Note that if your weights are sampling weights, then only the procedures whose names start with SURVEY (such as SURVEYLOGISTIC) have the necessary variance estimators to provide a proper analysis. The WEIGHT statement in PROC LOGISTIC is not appropriate for modeling survey data. That being said, the OUTROC= option provides sensitivities and specificities. You can use the ROCOPTIONS(WEIGHTED) option in the PROC LOGISTIC statement to have the weights multiply the counts as shown in the "Details: Receiver Operating Characteristic Curves: ROC Computations" section of the LOGISTIC documentation.
... View more
08-30-2012
01:14 PM
If you treat this as a problem of testing agreement among 5 raters, you can use the multiple rater kappa provided by the MAGREE macro: http://support.sas.com/kb/25006
... View more
08-30-2012
01:08 PM
Unfortunately, the ODDSRATIO statement did not honor the REF= level set by the CLASS statement until the current (SAS 9.3 TS1M2) release. Prior to that release, you canuse the LSMEANS statement with the DIFF=CONTROL, EXP, and CL options instead of the ODDSRATIO statement. The LSMEANS statement requires that you also use the PARAM=GLM option in the CLASS statement. For example, if variable A has values 1 and 2 and you want the odds ratio with the odds for level 1 in the denominator, the following LSMEANS statement provides the odds ratio estimate and confidence limits in the "Exponentiated" columns of the "Differences of A Least Squares Means" table. proc logistic; class a b / param=glm; model y=a b; lsmeans a / diff=control('2') exp cl; run;
... View more
08-30-2012
12:53 PM
You can use PROC GENMOD to estimate the poisson mean (lambda). For example, using the portion of data you provided: data a; input LOS Code; datalines; 4 500 3 500 7 501 2 200 8 501 10 501 12 501 ; proc genmod; model los = / dist=poisson; estimate 'lambda' intercept 1; run; The intercept reported (1.8827) is the estimated log mean. So, exp(1.8827)=6.57 is the estimated poisson mean. Assuming you'll want a confidence interval, the ESTIMATE statement merely repeats the estimate of the intercept which provides the mean estimate and confidence limits (4.92, 8.77). A histogram of the observed data can be done using the HISTOGRAM statement in PROC UNIVARIATE: proc univariate noprint; var los; hist los; run;
... View more