Formatting Issue in dtable: How to Collapse Binary Variables into a Compact Table?

[Tim Wallner]

Hello everyone,

I am using dtable in Stata to summarize binary variables (hasSepsis, hasInfection, hasUTI, hasOther) by a treatment variable (treated). However, my table output is not quite how I would like.

I would like to have the results on one row with the corresponding variable name.

I know how to hide the rows with the variable names and only show the rows with the results, but i'd rather not have to manually re-label all the "1"s to something meaningful. Rather I'm looking for a way to shift the results up to the variable name.

Current output:

--------------------------------------------
              Control   Treated      All    
--------------------------------------------
N            3 (30.0%) 7 (70.0%) 10 (100.0%)
hasSepsis                                   
  1          1 (33.3%)  0 (0.0%)   1 (10.0%)
hasInfection                                
  1          2 (66.7%)  0 (0.0%)   2 (20.0%)
hasUTI                                      
  1           0 (0.0%) 1 (14.3%)   1 (10.0%)
hasOther                                    
  1           0 (0.0%)  0 (0.0%)    0 (0.0%)
--------------------------------------------

Desired output, with results on one row with the variable name:

-------------------------------------------------
              Control    Treated     All    
-------------------------------------------------
N             3 (30.0%)  7 (70.0%)   10 (100.0%)
hasSepsis     1 (33.3%)  0 (0.0%)     1 (10.0%)
hasInfection  2 (66.7%)  0 (0.0%)     2 (20.0%)
hasUTI        0 (0.0%)   1 (14.3%)    1 (10.0%)
hasOther      0 (0.0%)   0 (0.0%)     0 (0.0%)
-------------------------------------------------

My code:

clear
input float treated byte(hasSepsis hasInfection hasUTI hasOther)
1 0 0 0 0
0 1 1 0 0
1 0 0 0 0
0 0 1 0 0
1 0 0 1 0
0 0 0 0 0
1 0 0 0 0
1 0 0 0 0
1 0 0 0 0
1 0 0 0 0
end

label values treated treated_lab
label def treated_lab 0 "Control" 1 "Treated", modify

dtable 1.has*, ///
by(treated) ///
column(by(hide) total(All) test(P-value))

[Hemanshu Kumar]

After your dtable command, do the following:

collect style header hasSepsis hasInfection hasUTI hasOther, level(hide)
collect preview

[Tim Wallner]

Thank you Hemanshu Kumar!

This gave me the wanted output:

--------------------------------------------
              Control   Treated      All    
--------------------------------------------
N            3 (30.0%) 7 (70.0%) 10 (100.0%)
hasSepsis    1 (33.3%)  0 (0.0%)   1 (10.0%)
hasInfection 2 (66.7%)  0 (0.0%)   2 (20.0%)
hasUTI        0 (0.0%) 1 (14.3%)   1 (10.0%)
hasOther      0 (0.0%)  0 (0.0%)    0 (0.0%)
--------------------------------------------

[Hemanshu Kumar]

If you want to fix the alignment issue that is causing the table in #3 to look a little different from the one in #1, you can additionally throw in

collect style cell, halign(left)

so that you get

. collect preview

--------------------------------------------
             Control   Treated   All        
--------------------------------------------
N            3 (30.0%) 7 (70.0%) 10 (100.0%)
hasSepsis    1 (33.3%) 0 (0.0%)  1 (10.0%)  
hasInfection 2 (66.7%) 0 (0.0%)  2 (20.0%)  
hasUTI       0 (0.0%)  1 (14.3%) 1 (10.0%)  
hasOther     0 (0.0%)  0 (0.0%)  0 (0.0%)   
--------------------------------------------