@RRzio11 wrote: Hello   I am trying to proc tabulate by age groups. I formatted my dataset by the following groups:   proc format; value age_cat 20-39="20>39" 40-49="40>49" 50-59="50>59" 60-200= ">60" .="Missing age" ;   and then put this as my proc tabulate code: PROC SORT DATA=NHM1514; BY AGE; RUN; proc tabulate data=nhm1514; CLASS AGE; var GENDER smoking martial education diabetes blood_pressure hiv bmi crp SHBG EST TST WBC ORXH06 ORXH11 ORXH16 ORXH18 ORXH26 ORXH31 ORXH33 ORXH35 ORXH39 ORXH40 ORXH42 ORXH45 ORXH51 ORXH52 ORXH53 ORXH54 ORXH55 ORXH56 ORXH58 ORXH59 ORXH61 ORXH62 ORXH64 ORXH66 ORXH67 ORXH68 ORXH69 ORXH70 ORXH71 ORXH72 ORXH73 ORXH81 ORXH82 ORXH83 ORXH84 LBDR06 LBDR11 LBDR16 LBDR18 LBDR26 LBDR31 LBDR33 LBDR35 LBDR39 LBDR40 LBDR42 LBDR45 LBDR51 LBDR52 LBDR53 LBDR54 LBDR55 LBDR56 LBDR58 LBDR59 LBDR61 LBDR62 LBDR64 LBDR66 LBDR67 LBDR68 LBDR69 LBDR70 LBDR71 LBDR72 LBDR73 LBDR81 LBDR82 LBDR83 LBDR84 LBDR89 HPVP06 HPVP11 HPVP16 HPVP18 HPVP26 HPVP31 HPVP33 HPVP35 HPVP39 HPVP40 HPVP42 HPVP45 HPVP51 HPVP52 HPVP53 HPVP54 HPVP55 HPVP56 HPVP58 HPVP59 HPVP61 HPVP62 HPVP64 HPVP66 HPVP67 HPVP68 HPVP69 HPVP70 HPVP71 HPVP72 HPVP73 HPVP81 HPVP82 HPVP83 HPVP84 HPVP89; table ( GENDER smoking martial education diabetes blood_pressure hiv bmi SHBG EST TST WBC ORXH06 ORXH11 ORXH16 ORXH18 ORXH26 ORXH31 ORXH33 ORXH35 ORXH39 ORXH40 ORXH42 ORXH45 ORXH51 ORXH52 ORXH53 ORXH54 ORXH55 ORXH56 ORXH58 ORXH59 ORXH61 ORXH62 ORXH64 ORXH66 ORXH67 ORXH68 ORXH69 ORXH70 ORXH71 ORXH72 ORXH73 ORXH81 ORXH82 ORXH83 ORXH84 LBDR06 LBDR11 LBDR16 LBDR18 LBDR26 LBDR31 LBDR33 LBDR35 LBDR39 LBDR40 LBDR42 LBDR45 LBDR51 LBDR52 LBDR53 LBDR54 LBDR55 LBDR56 LBDR58 LBDR59 LBDR61 LBDR62 LBDR64 LBDR66 LBDR67 LBDR68 LBDR69 LBDR70 LBDR71 LBDR72 LBDR73 LBDR81 LBDR82 LBDR83 LBDR84 LBDR89 HPVP06 HPVP11 HPVP16 HPVP18 HPVP26 HPVP31 HPVP33 HPVP35 HPVP39 HPVP40 HPVP42 HPVP45 HPVP51 HPVP52 HPVP53 HPVP54 HPVP55 HPVP56 HPVP58 HPVP59 HPVP61 HPVP62 HPVP64 HPVP66 HPVP67 HPVP68 HPVP69 HPVP70 HPVP71 HPVP72 HPVP73 HPVP81 HPVP82 HPVP83 HPVP84 HPVP89)*(N MEAN STD pctn), gender ALL; run;   but then i get 496 error messages on my log: ERROR: There are multiple analysis variables associated with a single table cell in the following nesting : GENDER * N * GENDER. ERROR: There are multiple analysis variables associated with a single table cell in the following nesting : GENDER * Mean * GENDER. ERROR: There are multiple analysis variables associated with a single table cell in the following nesting : GENDER * Std * GENDER. ERROR: There are multiple analysis variables associated with a single table cell in the following nesting : GENDER * PctN * GENDER. ERROR: There are multiple analysis variables associated with a single table cell in the following nesting : SMOKING * N * GENDER.   Can you help me understand what I am doing wrong and how to appropriately code by variable groups?   In the code you show the only variable that can have the role of a "group" variable is AGE. Any variable on the VAR statement is an analysis variable and will have statistic calculated for it. However Proc Tabulate does not allow you to "cross" var varaibles against each other because each needs a statistic and the intent of which is desired is not clear in that case. If you want Gender as group variable then 1) remove it from the VAR statement, 2) place it in a CLASS statement, and 3) remove it from the parenthesized variables in the Table statement.   While VAR variables must be numeric you can use numeric variables as Class variables. A format such as you show is one of the nice ways to create usable categories from a variable with many levels.   I suspect that you missed a FORMAT age age_cat.; and from your sort you likely intended a BY Age; in proc tabulate as well.   I would suspect that marital and education should more likely be Class variables and if smoking, diabetes and hiv are indicators of history of smoking, have/do not have diabetes or hiv that they are more likely candidates as class variables.   ... View more

Everything you ask for is presented in one or another of the examples for the various procedures in the documentation.  Why not try some of them out? The code and representative data are available for every example.  If you get something that doesn't seem to make sense, or just breaks down completely, then you could come back and ask us how we might address those problems.  But I, for one, am not going in blindly and start making suggestions.   You have contributed to a consultant's nightmare - a client brings in data, says he would like a list of things, but never describes the question their data was supposed to address, or how the data were collected, or who their ultimate audience might be, or what decisions may be made based on the results of the analysis.   SteveDenham ... View more

SAS dates are counts of days, so you only need a simple subtraction. ... View more

Why do you create the libref if you are not going to use it?   To reference a dataset using the libref you need to use a "two level" name. A name with a period in the middle.  If you use just a "one level" name then SAS will assume you want to use WORK as the libref. So this statement data psych; is the same as  data work.psych; To store this original data into "permanent" dataset in the PSYCH libref (or "library") you would use something like: data psych.original;   ... View more