About ChuksManuel

ChuksManuel · ‎06-17-2023

Figured it out. Thanks

ChuksManuel · ‎06-17-2023

Hi programmers, I have a variable assessed in two points (longitudinal data): 2019 survey, FS (coded as 0,1). 2020 survey, FS (coded as 0,1). I want to create a change variable such that; An observation with 0 in 2019 who is still 0 in 2020= variable A An observation with 0 in 2019 who becomes 1 in 2020=variable B An observation with 1 in 2019 who becomes 0 in 2020= variable C An observation with 1 in 2019 who is still 1 in 2020=variable D Is there a way I can code this? Do I need to create a flag variable in both years?

ChuksManuel · ‎06-04-2023

Hi all, I am trying to calculate age and sex prevalence rate for hypertension by survey year. Is there a way I can include age and sex variables in proc surveyfreq procedure? proc surveyfreq data=final_control; weight LLCPWT_F; strata _ststr; cluster _psu; tables year*hypertension/row col cl chisq; run; Age and sex are categorical variables.

ChuksManuel · ‎05-17-2023

Thanks!

ChuksManuel · ‎05-17-2023

Hello, It's not giving me any error. It's however not giving me what i want. I want the new variable created "abuse" to have 0,1 and missing. With "abuse" called to missing if an observation has missing in all the 5 composite variables. I will appreciate if you can refer me to materials to help me figure this out. I know i'm just missing a small piece of code.

ChuksManuel · ‎05-17-2023

Can anyone help me review this code? I'm trying to create a composite variable (abuse) but also to create a situation where any missing value across the 5 variables is coded as missing in the new composite variable. if ace_swear =1 or ace_hurt=1 or ace_touch=1 or ace_tthem=1 or ace_sex=1 then abuse=1; if ace_swear =. and ace_hurt=. and ace_touch=. and ace_tthem=. and ace_sex=. then abuse=.; else abuse=0;

ChuksManuel · ‎03-02-2023

Thanks!

ChuksManuel · ‎03-02-2023

Hi SAS procedure community, I have a survey data that i want to calculate the relative risk. I know i can implement this in SAS proc genmod using dist=binomial and link=log option. However i want to account for the complex sample design (design weight, strata and cluster). I saw the %surveygenmod macro which is the sas version of svyglm in R. The macro isn't already in SAS. Can anyone direct me on how I can get this uploaded in my SAS environment for use? Thanks

ChuksManuel · ‎02-19-2023

The mistake was in the final else statement. Switched them and it worked. Thanks!

ChuksManuel · ‎02-19-2023

Hi programmers, I have a variable poorhlth (1-30 days). 88 in the coding represents "no days of poor health". And 77, 99 represent refused/don't know. I want to creat a new variable poor_count that will set 88 as zero, and treat 77 and 99 as missing. Below is my code. What did i do wrong here? if POORHLTH= '88' then poor_count=0; else if POORHLTH in (77,99) then poor_count=.; else POORHLTH= poor_count;

ChuksManuel · ‎02-06-2023

1. Can't i use a simple merge to execute this instead of proc sql? 2. There's no overlap in ID in this small dataset (obs=7 to 10) because the dataset is large. But the joining dataset has all the unique IDs (HHX_2019) for the 2019 data and HHX_2020 for the 2020 data. I renamed the HHX variables in the 2020 and 2019 datasets to HHX_2020 and HHX_2019 to correspond with the one in the ID in the joining datatset.

ChuksManuel · ‎02-06-2023

/*2019 Dataset has this structure*/ RECTYPE SRVY_YR HHX REGION PSTRAT PPSU 10 Sample Adult 2019 H048109 3 South 122 2 10 Sample Adult 2019 H027044 3 South 122 2 10 Sample Adult 2019 H058855 3 South 122 2 10 Sample Adult 2019 H031993 3 South 122 2 10 Sample Adult 2019 H007122 3 South 115 2 10 Sample Adult 2019 H007736 3 South 115 2 10 Sample Adult 2019 H040698 3 South 115 1 10 Sample Adult 2019 H022161 3 South 115 1 10 Sample Adult 2019 H017054 3 South 115 1 /*2020 dataset has this structure*/ RECTYPE SRVY_YR HHX REGION PSTRAT PPSU 10 2020 H066706 3 Almost everything 103 2 10 2020 H034928 3 Almost everything 103 2 10 2020 H018289 3 Almost everything 103 2 10 2020 H006876 3 Almost everything 103 2 10 2020 H028842 3 Almost everything 103 2 10 2020 H004811 3 Almost everything 103 2 10 2020 H068043 3 Almost everything 103 2 /*joining dataset has this structure*/ RECTYPE SRVY_YR HHX_2019 HHX_2020 WTSA_L 60 Sample Adult Longitudinal Survey year H000003 H038763 33091.61 60 Sample Adult Longitudinal Survey year H000008 H060250 15476.27 60 Sample Adult Longitudinal Survey year H000009 H018743 15994.46 60 Sample Adult Longitudinal Survey year H000010 H067593 13108.81 60 Sample Adult Longitudinal Survey year H000011 H021952 8211.29 60 Sample Adult Longitudinal Survey year H000012 H007892 40972.78 60 Sample Adult Longitudinal Survey year H000014 H041303 16721.16 60 Sample Adult Longitudinal Survey year H000023 H004538 1852.21 60 Sample Adult Longitudinal Survey year H000027 H007281 4790.91

ChuksManuel · ‎02-05-2023

And for the files, I have selected just the first 20 observations

ChuksManuel · ‎02-05-2023

Hello programmers, I am want to do an analysis to estimate the risk of developing stroke in those who have a heart disease in the recently released NHIS longitudinal dataset. The NHIS released a longitudinal dataset last year that makes it possible for researchers to merge the 2019 and 2020 together. So my plan is to have a sample of people with heart disease (after applying all exclusion criteria) in the 2019 dataset and check to see if they have the outcome (stroke) in the 2020 dataset. I intend to use cox regression while controlling for baseline covariates. I have however not been able to merge the 2019 dataset and the 2020 dataset with the joining file. I will appreciate if anyone can help me merge it in a way for me to conduct my analysis. The 2019 dataset has over 300k obs, 2020 has 31k and the join dataset (people followed up from 2019 to 2020) is 10k. My initial code is as follows but i know I'm missing something. I have attached the 2019, 2020 and join dataset in excel as well. data merged_data; merge nhis.adultlong20 (in=a) one (in=b) two(in=c); by HHX_2019 ; /* the common variable */ run;

ChuksManuel · ‎01-12-2023

Hello programmers, I'll appreciate if anyone can give me an insight. I am trying to assign a value to a cluster of individuals. Each individual have a yes or no response and i want to assign the proportion of yeses to all in the individual in that cluster. Eg. ID is nested in cluster 104; responses are yeses and nos. value is the proportion of "yeses"= 6/20=0.3. My challenge is how to call this value on every ID in the cluster. Any help is appreciated. ID cluster Ab value 1 104 yes 0.3 2 104 no 0.3 3 104 no 0.3 4 104 no 0.3 5 104 no 0.3 6 104 no 0.3 7 104 no 0.3 8 104 yes 0.3 9 104 yes 0.3 10 104 yes 0.3 11 104 yes 0.3 12 104 yes 0.3 13 104 no 0.3 14 104 no 0.3 15 104 no 0.3 16 104 no 0.3 17 104 no 0.3 18 104 no 0.3 19 104 no 0.3 20 104 no 0.3

Online Status	Offline
Date Last Visited	‎06-17-2023 02:15 PM

Re: Indicating change in a variable across a year

Indicating change in a variable across a year

age and sex adjusted prevalence rate

Re: if and then statement

Re: if and then statement

if and then statement

Re: Complex sample analysis with proc genmod

Complex sample analysis with proc genmod

Re: If and else statement

If and else statement

Re: Creating a variable from other variables

Re: Creating a variable from other variables

Re: Logistic Regression for trend analysis using Complex survey data?

Re: Logistic Regression for trend analysis using Complex survey data?

Re: Logistic Regression for trend analysis using Complex survey data?

Re: Does anyone know where i can download new versions of NLMeans 1.3 ...

Re: Indicating change in a variable across a year

Indicating change in a variable across a year

age and sex adjusted prevalence rate

Re: if and then statement

Re: if and then statement

if and then statement

Re: Complex sample analysis with proc genmod

Complex sample analysis with proc genmod

Re: If and else statement

If and else statement

Re: file merge

Re: file merge

Re: file merge

file merge

How to assign a proportion to a cluster in multilevel