cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 
quentinread
‎08-15-2024
  • 5 Posts
  • 7 Likes Given
  • 1 Solutions
  • 2 Likes Received

Re: How to format estimate statement to get a mean across non-empty ce...

by Fluorite | Level 6 in Statistical Procedures
‎08-08-2024 07:36 AM
2 Likes
‎08-08-2024 07:36 AM
2 Likes
Thank you for this very helpful comment! It got me 99% of the way there. The bylevel option did not give me exactly what I wanted because I wanted the lsmean of each species only across the months 7-9 cells. Using bylevel gives the correct lsmean for species 2 and 3 but not species 1. But your suggestion to use OM inspired me to look in the documentation. I found that creating a new dataset for obsmargins that does not include month 6, and passing that as the obsmargins dataset in proc plm, I can get the correct lsmeans that I was looking for, see code below.   data nojune; set example; where month <> '6'; run; proc glimmix data=example; class species month pasture block; model biomass=species|month / dist=gamma link=log; random pasture(block); store out=examplemod; run; proc plm restore=examplemod; lsmeans species / cl obsmargins=nojune pdiff adjust=tukey; run; ... View more

How to format estimate statement to get a mean across non-empty cells

by Fluorite | Level 6 in Statistical Procedures
‎08-07-2024 11:33 AM
‎08-07-2024 11:33 AM
Hello, I know there are similar questions on this forum but being a SAS noob, I have been unable to get anything to work for my particular use case. Below I provide a simple example dataset that illustrates the issue. I have an "unbalanced" dataset where I have 3 species, sp1, sp2, and sp3, for which biomass was measured. But sp1 was measured in months 6,7,8,9 while sp2 and sp3 were not measured in month 6 but only in 7,8,9.   I understand why the lsmeans statement returns non-estimable for species 2 and 3 because you cannot get a least-square mean over empty cells. However I want to write an estimate statement to get means for sp2 and sp3, only over months 7,8,9 and not considering month 6. I still get non-estimable in that case. I don't understand whether I am writing the statement incorrectly or whether I am misunderstanding what the estimate statement is doing.   Thanks for your help!   DATA example; INPUT pasture $ block $ month $ year $ species $ biomass; cards; 3 1 6 2021 sp1 745 6 2 6 2021 sp1 1011 9 3 6 2021 sp1 1129 3 1 7 2021 sp1 1612 6 2 7 2021 sp1 2034 9 3 7 2021 sp1 2210 3 1 8 2021 sp1 2795 6 2 8 2021 sp1 2975 9 3 8 2021 sp1 2699 3 1 9 2021 sp1 499 6 2 9 2021 sp1 624 9 3 9 2021 sp1 759 1 1 7 2021 sp2 2422 2 1 7 2021 sp3 2685 4 2 7 2021 sp3 1276 5 2 7 2021 sp2 2303 7 3 7 2021 sp2 2269 8 3 7 2021 sp3 784 1 1 8 2021 sp2 6833 2 1 8 2021 sp3 3604 4 2 8 2021 sp3 2231 5 2 8 2021 sp2 8447 7 3 8 2021 sp2 6303 8 3 8 2021 sp3 1656 1 1 9 2021 sp2 3675 2 1 9 2021 sp3 1129 4 2 9 2021 sp3 1552 5 2 9 2021 sp2 6495 7 3 9 2021 sp2 5575 8 3 9 2021 sp3 647 ; proc glimmix data=example; class species month pasture block; model biomass=species|month / e dist=gamma link=log s; random pasture(block); lsmeans species; * Species 2 and 3 are not estimable which makes sense because they have no data for month 6; estimate "Species 2, only months 7-9" species 0 1 0 month 0 1 1 1; * Also not estimable but I don't understand why not because I am not averaging over any empty cells; run; ... View more

Re: Relative risks between multinomial categories within a population ...

by Fluorite | Level 6 in Statistical Procedures
‎09-26-2022 09:22 AM
‎09-26-2022 09:22 AM
Thanks @StatDave , that is definitely what I was looking for. I hate to keep asking questions on this same thread, but do you have any idea how to use Hans-Peter Piepho's mult macro to get the multiple comparison letters for those adjusted p-values? I have tried to use it but it gives me nonsensical output.   Here is what I have tried: first, label the comparisons by interval so that the proc multtest does the Holm adjustment within each interval.   data my_nlmeans; set est; interval=ceil(_n_/6); run; proc multtest inpvalues(pr)=my_nlmeans holm; by interval; id label; ods output pvalues=mult_pvals; run; Next, manually label the p-value data object and the nlmeans data object with columns for the treatments being compared, as I believe is required by the mult macro. Then run the mult macro for an individual interval. This gives me "a" for every group even though some of the adjusted p-values are below 0.05. Maybe I am using it incorrectly?   data tissue_contrast_labels; infile datalines missover; input interval _interval response $10. _response $10.; datalines; 1 1 berries female 1 1 berries leaftip 1 1 berries male 1 1 female leaftip 1 1 female male 1 1 leaftip male 2 2 berries female 2 2 berries leaftip 2 2 berries male 2 2 female leaftip 2 2 female male 2 2 leaftip male 3 3 berries female 3 3 berries leaftip 3 3 berries male 3 3 female leaftip 3 3 female male 3 3 leaftip male 4 4 berries female 4 4 berries leaftip 4 4 berries male 4 4 female leaftip 4 4 female male 4 4 leaftip male ; run; data mult_pvals; merge mult_pvals tissue_contrast_labels; run; data my_nlmeans; merge my_nlmeans tissue_contrast_labels; run; %mult(diffs=mult_pvals, lsmeans=my_nlmeans, trt=response, p=stepdownbonferroni, by=interval, level=1); ... View more

Re: Relative risks between multinomial categories within a population ...

by Fluorite | Level 6 in Statistical Procedures
‎09-22-2022 01:36 PM
‎09-22-2022 01:36 PM
Hi Koen, yes, thanks very much for your response. I am hacking away at the NLMeans macro right now (don't worry I did not try to edit the source code, haha), and I think I am close to the correct solution. I believe I have all the correct contrasts. I have one other question: is it possible within the NLMeans macro to automatically adjust the p-values for multiple comparisons, or must I do this manually by calculating my desired adjustment to alpha and inputting that value to the alpha argument?   This is what I have so far:   proc gee data=example; class response(ref="No response") trial rep interval; model response=interval / dist=multinomial link=glogit; repeated subject=rep(trial); lsmeans interval / e diff ilink oddsratio; ods output lsmeans=gee1_lsmeans coef=gee1_coeffs; store out=gee1_mod; run; data tissue_contrasts; infile datalines missover; input label $40. k1-k20; set=1; datalines; Interval 1 P(leaftip)/P(male) 0 0 0 0 0 0 0 0 1 0 0 0 -1 0 0 0 0 0 0 0 Interval 1 P(leaftip)/P(female) 0 0 0 0 -1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 Interval 1 P(leaftip)/P(berries) -1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 Interval 1 P(male)/P(female) 0 0 0 0 -1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 Interval 1 P(male)/P(berries) -1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 Interval 1 P(female)/P(berries) -1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Interval 2 P(leaftip)/P(male) 0 0 0 0 0 0 0 0 0 1 0 0 0 -1 0 0 0 0 0 0 Interval 2 P(leaftip)/P(female) 0 0 0 0 0 -1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 Interval 2 P(leaftip)/P(berries) 0 -1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 Interval 2 P(male)/P(female) 0 0 0 0 0 -1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 Interval 2 P(male)/P(berries) 0 -1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 Interval 2 P(female)/P(berries) 0 -1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Interval 3 P(leaftip)/P(male) 0 0 0 0 0 0 0 0 0 0 1 0 0 0 -1 0 0 0 0 0 Interval 3 P(leaftip)/P(female) 0 0 0 0 0 0 -1 0 0 0 1 0 0 0 0 0 0 0 0 0 Interval 3 P(leaftip)/P(berries) 0 0 -1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 Interval 3 P(male)/P(female) 0 0 0 0 0 0 -1 0 0 0 0 0 0 0 1 0 0 0 0 0 Interval 3 P(male)/P(berries) 0 0 -1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 Interval 4 P(female)/P(berries) 0 0 -1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 Interval 4 P(leaftip)/P(male) 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 -1 0 0 0 0 Interval 4 P(leaftip)/P(female) 0 0 0 0 0 0 0 -1 0 0 0 1 0 0 0 0 0 0 0 0 Interval 4 P(leaftip)/P(berries) 0 0 0 -1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 Interval 4 P(male)/P(female) 0 0 0 0 0 0 0 -1 0 0 0 0 0 0 0 1 0 0 0 0 Interval 4 P(male)/P(berries) 0 0 0 -1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 Interval 4 P(female)/P(berries) 0 0 0 -1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 ; run; %NLMeans(instore=gee1_mod, coef=gee1_coeffs, link=glogit, options=ratio, contrasts=tissue_contrasts, null=1, title=Contrasts between tissue types within each interval);   ... View more

Relative risks between multinomial categories within a population in p...

by Fluorite | Level 6 in Statistical Procedures
‎09-22-2022 09:29 AM
‎09-22-2022 09:29 AM
Disclaimer: I am really an R user but helping a client with their SAS code. I understand the basics of SAS but have trouble with customizing output to give me exactly the tests I want, hence this question!   I have a model I fit with proc gee that estimates the probability of insects being attracted to various types of plant tissue as a function of time. At any given time, an insect may choose between one of four tissue types or not respond at all. There were 4 trials, 10 replicates within each trial, and each replicate consisted of 20 insects. Within each replicate, at 4 time points, the number of insects on each of the tissue types was counted. So we have repeated measures data with a non-ordered multinomial outcome. I chose to use proc gee to fit this model. Following is the code, with a small subset of the data, enough to provide a reproducible example. Each row in the data is an individual insect at a single time point (interval), with the character column "response" having five possible values: Leaftip, Male, Female, Berries, and No response.   I was able to get the contrasts between different levels of "interval" (the time point) using the lsmeans statement, but I do not know how to get contrasts and the appropriate statistical test, to compare (1) the relative preference (relative risk) between the response categories within each interval and (2) the relative preference between the response categories, averaged across all four intervals. The hypothesis of interest to test would be, for example, "do insects have a higher preference for leaf tips relative to berries at time interval 1?" I found this note on calculating relative risks in a multinomial response model which seems to be doing something similar but I am not sure how to adapt the macro to work for this particular case. The examples given are for different procedures, not proc gee. Any help would be greatly appreciated!     data example; input trial rep interval response $12.; datalines; 1 1 1 Leaftip 1 1 1 Leaftip 1 1 1 Leaftip 1 1 1 Leaftip 1 1 1 Male 1 1 1 Male 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 1 No response 1 1 2 Leaftip 1 1 2 Leaftip 1 1 2 Leaftip 1 1 2 Leaftip 1 1 2 Leaftip 1 1 2 Leaftip 1 1 2 Leaftip 1 1 2 Leaftip 1 1 2 Leaftip 1 1 2 Male 1 1 2 Male 1 1 2 No response 1 1 2 No response 1 1 2 No response 1 1 2 No response 1 1 2 No response 1 1 2 No response 1 1 2 No response 1 1 2 No response 1 1 2 No response 1 1 3 Leaftip 1 1 3 Leaftip 1 1 3 Leaftip 1 1 3 Leaftip 1 1 3 Leaftip 1 1 3 Leaftip 1 1 3 Leaftip 1 1 3 Leaftip 1 1 3 Leaftip 1 1 3 Leaftip 1 1 3 Male 1 1 3 Male 1 1 3 Berries 1 1 3 No response 1 1 3 No response 1 1 3 No response 1 1 3 No response 1 1 3 No response 1 1 3 No response 1 1 3 No response 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 Leaftip 1 1 4 No response 1 1 4 No response 1 1 4 No response 1 1 4 No response 1 2 1 Leaftip 1 2 1 Leaftip 1 2 1 Leaftip 1 2 1 Leaftip 1 2 1 Male 1 2 1 Female 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 1 No response 1 2 2 Leaftip 1 2 2 Leaftip 1 2 2 Leaftip 1 2 2 Leaftip 1 2 2 Leaftip 1 2 2 Leaftip 1 2 2 Berries 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 2 No response 1 2 3 Leaftip 1 2 3 Leaftip 1 2 3 Leaftip 1 2 3 Leaftip 1 2 3 Leaftip 1 2 3 Male 1 2 3 Berries 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 3 No response 1 2 4 Leaftip 1 2 4 Leaftip 1 2 4 Leaftip 1 2 4 Leaftip 1 2 4 Leaftip 1 2 4 Leaftip 1 2 4 Leaftip 1 2 4 Male 1 2 4 Male 1 2 4 No response 1 2 4 No response 1 2 4 No response 1 2 4 No response 1 2 4 No response 1 2 4 No response 1 2 4 No response 1 2 4 No response 1 2 4 No response 1 2 4 No response 1 2 4 No response 2 1 1 Leaftip 2 1 1 Leaftip 2 1 1 Leaftip 2 1 1 Berries 2 1 1 Berries 2 1 1 Berries 2 1 1 Berries 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 1 No response 2 1 2 Leaftip 2 1 2 Leaftip 2 1 2 Leaftip 2 1 2 Leaftip 2 1 2 Leaftip 2 1 2 Leaftip 2 1 2 Leaftip 2 1 2 Berries 2 1 2 Berries 2 1 2 Berries 2 1 2 Berries 2 1 2 No response 2 1 2 No response 2 1 2 No response 2 1 2 No response 2 1 2 No response 2 1 2 No response 2 1 2 No response 2 1 2 No response 2 1 2 No response 2 1 3 Leaftip 2 1 3 Leaftip 2 1 3 Leaftip 2 1 3 Leaftip 2 1 3 Leaftip 2 1 3 Leaftip 2 1 3 Leaftip 2 1 3 Berries 2 1 3 Berries 2 1 3 No response 2 1 3 No response 2 1 3 No response 2 1 3 No response 2 1 3 No response 2 1 3 No response 2 1 3 No response 2 1 3 No response 2 1 3 No response 2 1 3 No response 2 1 3 No response 2 1 4 Leaftip 2 1 4 Leaftip 2 1 4 Leaftip 2 1 4 Leaftip 2 1 4 Leaftip 2 1 4 Leaftip 2 1 4 Leaftip 2 1 4 Male 2 1 4 Female 2 1 4 Female 2 1 4 Berries 2 1 4 Berries 2 1 4 Berries 2 1 4 No response 2 1 4 No response 2 1 4 No response 2 1 4 No response 2 1 4 No response 2 1 4 No response 2 1 4 No response 2 2 1 Leaftip 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 1 No response 2 2 2 Leaftip 2 2 2 Leaftip 2 2 2 Leaftip 2 2 2 Leaftip 2 2 2 Male 2 2 2 Female 2 2 2 Female 2 2 2 Female 2 2 2 Female 2 2 2 Female 2 2 2 Berries 2 2 2 No response 2 2 2 No response 2 2 2 No response 2 2 2 No response 2 2 2 No response 2 2 2 No response 2 2 2 No response 2 2 2 No response 2 2 2 No response 2 2 3 Leaftip 2 2 3 Leaftip 2 2 3 Male 2 2 3 Female 2 2 3 Female 2 2 3 Female 2 2 3 Female 2 2 3 Female 2 2 3 No response 2 2 3 No response 2 2 3 No response 2 2 3 No response 2 2 3 No response 2 2 3 No response 2 2 3 No response 2 2 3 No response 2 2 3 No response 2 2 3 No response 2 2 3 No response 2 2 3 No response 2 2 4 Leaftip 2 2 4 Leaftip 2 2 4 Leaftip 2 2 4 Leaftip 2 2 4 Leaftip 2 2 4 Leaftip 2 2 4 Leaftip 2 2 4 Male 2 2 4 Male 2 2 4 Female 2 2 4 Berries 2 2 4 Berries 2 2 4 Berries 2 2 4 No response 2 2 4 No response 2 2 4 No response 2 2 4 No response 2 2 4 No response 2 2 4 No response 2 2 4 No response ; run; proc gee data=example; class response trial rep interval; model response=interval / dist=multinomial link=glogit; repeated subject=rep(trial); lsmeans interval / diff ilink oddsratio; run;   ... View more
Contact Me
Online Status
Offline
Date Last Visited
‎08-15-2024 03:46 PM
Likes given to
Likes from
View All
Quick links