Combining collections of results

[Kehinde Atoloye]

I am trying to combine different results in different collections into one but am unable to figure out how to go about it. Perhaps I'm going about it wrongly. I want the combined results to look like the sample below. I would like to be able to use collect export to export to excel at the end. Kindly help. Thank you.




Below is my data and my code so far.

* Example generated by -dataex-. For more info, type help dataex
clear
input float q408a_binary byte(exposure period q104cat) float(wealth_quintile iptw_stabilized) str4 clustercode
1 1 1 2 1         . "2128"
1 1 1 2 3  .6334819 "2128"
1 1 1 1 2  .8872965 "2128"
1 1 1 2 1   1.63046 "2128"
1 1 1 2 1         . "2128"
1 1 1 2 2 1.2247086 "2128"
1 1 1 2 1         . "2128"
1 1 1 1 2  1.012563 "2128"
1 1 1 1 3  .6614283 "2128"
1 1 1 1 2 1.0122956 "2128"
1 1 1 2 2 1.2247086 "2128"
1 1 1 1 1 1.8715497 "2128"
0 1 1 2 3  .6759744 "2129"
1 1 1 2 2 1.0476002 "2129"
1 1 1 1 1 1.1034142 "2129"
1 1 1 2 2 1.0593003 "2129"
1 1 1 1 2  .8957604 "2129"
1 1 1 2 1         . "2129"
1 1 1 2 3         . "2129"
1 1 1 1 2  .8955543 "2129"
1 1 1 2 4  .5979365 "2129"
1 1 1 1 3  .5950213 "2129"
1 1 1 2 2 1.0593003 "2129"
1 1 1 2 1  1.606752 "2129"
1 1 1 2 2  .9227699 "2130"
1 1 1 1 1 1.2819333 "2130"
1 1 1 1 1 1.2990835 "2130"
1 1 1 2 1  1.606752 "2130"
1 1 1 1 1 1.2819333 "2130"
1 1 1 2 2  .9227699 "2130"
1 1 1 1 1 1.8715497 "2130"
1 1 1 1 1 1.5357575 "2130"
1 1 1 2 4  .5745883 "2130"
1 1 1 1 1 1.5357575 "2130"
1 1 1 1 2  1.012563 "2130"
1 1 1 2 2 1.0593003 "2130"
1 1 1 2 5  .5580806 "2131"
1 1 1 2 5  .5669054 "2131"
1 1 1 2 5  .5669054 "2131"
1 1 1 2 4  .5831169 "2131"
1 1 1 2 5  .5669054 "2131"
1 1 1 2 5  .5522093 "2131"
1 1 1 1 3  .6614283 "2131"
1 1 1 2 5  .5669054 "2131"
1 1 1 2 3   .672333 "2131"
1 1 1 2 5  .5669054 "2131"
1 1 1 1 5 .54803306 "2131"
1 1 1 2 5  .5522093 "2131"
1 1 1 2 3  .6334819 "2133"
0 1 1 2 3  .7324004 "2133"
1 1 1 2 5  .5669054 "2133"
1 1 1 2 4  .5959356 "2133"
1 1 1 2 3   .672333 "2133"
1 1 1 1 3         . "2133"
1 1 1 2 5  .5522093 "2133"
1 1 1 2 5  .5522093 "2133"
1 1 1 2 4  .5959356 "2133"
1 1 1 2 5  .5669054 "2133"
1 1 1 2 3  .7274547 "2133"
1 1 1 2 3         . "2133"
1 1 1 1 5 .54803306 "2134"
1 1 1 2 5  .5669054 "2134"
1 1 1 2 2 1.0593003 "2134"
1 1 1 1 4  .5899439 "2134"
1 1 1 2 3         . "2134"
1 1 1 2 2 1.0593003 "2134"
1 1 1 2 2 1.0476002 "2134"
1 1 1 2 3         . "2134"
1 1 1 2 3         . "2134"
1 1 1 2 4  .5959356 "2134"
1 1 1 2 5  .5669054 "2134"
1 1 1 2 5  .5878007 "2134"
1 1 1 2 5  .5878007 "2135"
0 1 1 1 5         . "2135"
1 1 1 2 5  .5580806 "2135"
1 1 1 2 5  .5669054 "2135"
1 1 1 2 4         . "2135"
1 1 1 1 4  .6182836 "2135"
1 1 1 2 4  .5979365 "2135"
1 1 1 2 4  .6262231 "2135"
1 1 1 2 4  .5745883 "2135"
0 1 1 1 5 .54803306 "2135"
1 1 1 2 5 .56828284 "2135"
1 1 1 2 4  .5959356 "2135"
1 1 1 2 4  .5745883 "2136"
1 1 1 2 5  .5669054 "2136"
1 1 1 2 3         . "2136"
1 1 1 2 5  .5669054 "2136"
1 1 1 2 4  .5959356 "2136"
0 1 1 2 5  .5437789 "2136"
1 1 1 2 5  .5877561 "2136"
1 1 1 2 5         . "2136"
1 1 1 2 5         . "2136"
1 1 1 2 5  .5669054 "2136"
1 1 1 2 4  .6262231 "2136"
0 1 1 2 5  .5669054 "2136"
1 1 1 2 5 .56828284 "2137"
1 1 1 2 4  .5959356 "2137"
1 1 1 2 5 .54540646 "2137"
1 1 1 2 5  .5580806 "2137"
end
label values q408a_binary q408a_binary_lbl
label def q408a_binary_lbl 0 "No", modify
label def q408a_binary_lbl 1 "Yes", modify
label values exposure exposure
label def exposure 1 "Intervention", modify
label values period period
label def period 1 "Endline", modify
label values q104cat q104cat
label def q104cat 1 "15-17", modify
label def q104cat 2 "18-19", modify
label values wealth_quintile wealth_quintile
label def wealth_quintile 1 "Very poor", modify
label def wealth_quintile 2 "Poor", modify
label def wealth_quintile 3 "Averga", modify
label def wealth_quintile 4 "Rich", modify
label def wealth_quintile 5 "Very rich", modify

melogit q408a_binary i.exposure##i.period i.q104cat i.wealth_quintile  [pweight=iptw_stabilized] || clustercode:, vce(cluster clustercode) or
// margins i.exposure#i.period, contrast(nowald) over(wealth_quintile)
collect clear
collect create main, replace
collect stars _r_b 0.05 "*", attach(_r_b) shownote
quietly collect: margins i.exposure#i.period, contrast(nowald)
collect style cell result, nformat(%8.3f) 
collect composite define blu = _r_lb _r_ub, delimiter(,) trim replace
collect style cell result[blu], sformat("[%s]")
collect style cell result[_r_b], sformat("%s")
collect composite define bci = _r_b blu stars, trim replace
collect style header result, title(hide)
collect label levels result bci "ATT [95% CI]" 
collect layout (colname) (result[bci])
collect save, name(main) replace


local cov q104cat wealth_quintile
foreach v of  local cov {
collect create `v', replace
collect stars _r_b 0.05 "*", attach(_r_b) shownote
quietly collect: margins i.exposure#i.period, contrast(nowald) over(`v')
collect style cell result, nformat(%8.3f) 
collect composite define blu = _r_lb _r_ub, delimiter(,) trim replace
collect style cell result[blu], sformat("[%s]")
collect style cell result[_r_b], sformat("%s")
collect composite define bci = _r_b blu stars, trim replace
collect label levels result bci "ATT [95% CI]" 
collect style header result, title(hide)
collect style header `v', title(label)
collect layout (`v') (result[bci])
collect save, name(`v') replace
}

[Hemanshu Kumar]

Your code fails when I try to run it because your data extract has no variation in either exposure or period. Please provide an extract with which we can replicate your issue.

[Kehinde Atoloye]

Your code fails when I try to run it because your data extract has no variation in either exposure or period. Please provide an extract with which we can replicate your issue.

Here is a more representative sample of the original dataset. Thank you

* Example generated by -dataex-. For more info, type help dataex
clear
input float q408a_binary byte(exposure period q104cat) float(wealth_quintile iptw_stabilized) str4 clustercode
1 0 1 2 4         . "2266"
0 0 0 1 1  .7392449 "2220"
0 0 1 1 1  .8113364 "2222"
0 1 0 2 5  .6228189 "1104"
0 0 0 2 2  .8435132 "2217"
1 1 0 2 4  .6262879 "2105"
0 1 0 1 2         . "1111"
0 0 0 1 2         . "2214"
0 0 1 2 4  .7383448 "4201"
1 1 1 1 4 .58989817 "2102"
1 1 1 1 2  1.632937 "1109"
1 1 1 2 5  .5522093 "2142"
0 0 1 1 2         . "4217"
0 0 0 2 4         . "2261"
0 1 0 2 4  .5959356 "2135"
1 1 1 2 4         . "2159"
1 1 1 2 3   .672333 "2119"
1 0 0 2 1 .51715845 "2231"
1 1 1 2 4  .5959356 "2109"
1 1 1 2 4  .5959356 "2134"
1 1 1 2 3   .672333 "2108"
0 0 0 1 3  2.073654 "2245"
0 0 1 1 2         . "2243"
0 1 0 1 1         . "1122"
1 1 1 2 3   .672333 "2152"
1 1 0 1 5 .58229035 "2118"
1 0 1 1 1 .52606916 "2240"
1 1 1 2 3  .7274547 "2153"
0 0 1 1 2         . "2241"
0 1 0 2 1  1.606752 "2101"
0 0 1 1 2         . "2265"
1 1 0 1 5   .562749 "2153"
0 0 1 1 1  .6798754 "2260"
1 1 0 1 5  .5627805 "2134"
1 0 1 2 2  .6567904 "1201"
1 0 0 2 1         . "2267"
0 1 0 1 1         . "2132"
1 0 1 1 5  2.417476 "3216"
0 1 0 2 4         . "2130"
0 1 0 2 3  .7274547 "2120"
1 1 1 2 2 1.2247086 "2139"
0 0 0 1 4         . "1219"
1 1 1 2 3         . "1102"
0 0 0 1 1         . "2221"
0 0 1 1 3         . "2247"
1 1 1 2 4  .5959356 "2107"
0 1 0 2 4  .5959356 "2128"
1 1 1 2 3  .6334819 "2128"
1 1 1 2 3   .672333 "2125"
1 1 0 2 4  .5959356 "2154"
0 1 0 1 5  .5627805 "2139"
1 1 1 2 3  .6334819 "2168"
0 0 0 2 5         . "1205"
1 1 1 1 4 .58989817 "2101"
0 1 0 1 4  .6918721 "1112"
0 1 0 2 4  .6262231 "2164"
0 0 0 2 5  5.553016 "2228"
1 1 1 2 3  .6334819 "2154"
1 1 1 1 5  .5627805 "2114"
0 1 1 2 3         . "1117"
0 0 0 2 2  .9464296 "2244"
0 0 0 2 4         . "2228"
0 1 0 1 1  1.535216 "2130"
1 1 0 2 4  .6262231 "2163"
1 0 0 2 1  .7909834 "2210"
0 1 0 2 1         . "1120"
0 1 0 1 3         . "2101"
0 0 1 2 5  2.293048 "1221"
0 0 0 2 1 .51758623 "2252"
0 1 1 2 3  .9904094 "1106"
1 0 1 2 4 1.4302355 "3216"
0 0 1 2 1         . "2234"
1 1 1 2 5  .5669054 "2112"
0 1 0 2 2 1.0593003 "2135"
0 1 0 2 3         . "2132"
1 1 1 2 3  .7275726 "2144"
0 0 0 2 1 .51758623 "2238"
0 0 1 1 1         . "4211"
1 1 1 1 5 .54803306 "2113"
0 0 0 2 3         . "1221"
0 0 0 1 1  .7393718 "2203"
0 0 0 1 1         . "2264"
1 0 1 1 2         . "1215"
0 0 0 2 1  .7189214 "2212"
0 0 0 1 1         . "2202"
0 0 0 1 3 1.0451207 "1215"
0 1 0 2 4  .7707077 "1114"
0 1 0 2 2         . "1120"
1 0 0 2 2  .5384695 "2226"
1 1 1 2 5  .5669054 "2117"
1 1 1 2 2 1.0593003 "2134"
1 1 1 2 5  .5522093 "2157"
1 0 0 2 3  .6297176 "2226"
0 0 1 1 5  .8715369 "2237"
0 1 0 1 1         . "1113"
0 0 0 2 2         . "2222"
1 1 1 2 4  .5959356 "2104"
1 1 1 2 5  .5580806 "2122"
0 0 0 2 1         . "2268"
0 0 1 2 3         . "2226"
end
label values q408a_binary q408a_binary_lbl
label def q408a_binary_lbl 0 "No", modify
label def q408a_binary_lbl 1 "Yes", modify
label values exposure exposure
label def exposure 0 "Control", modify
label def exposure 1 "Intervention", modify
label values period period
label def period 0 "Baseline", modify
label def period 1 "Endline", modify
label values q104cat q104cat
label def q104cat 1 "15-17", modify
label def q104cat 2 "18-19", modify
label values wealth_quintile wealth_quintile
label def wealth_quintile 1 "Very poor", modify
label def wealth_quintile 2 "Poor", modify
label def wealth_quintile 3 "Averga", modify
label def wealth_quintile 4 "Rich", modify
label def wealth_quintile 5 "Very rich", modify

[Hemanshu Kumar]

You can add this code at the end of yours:

collect set main
collect addtags sections[main]
collect set q104cat
collect addtags sections[subpop]
collect set wealth_quintile
collect addtags sections[subpop]


collect combine comb = main q104cat wealth_quintile
collect layout (sections # (colname[1.exposure#1.period] q104cat wealth_quintile)) (result[bci])
collect label dim wealth_quintile "Wealth quintile"
collect label dim q104cat "Age (categories)"
collect label levels sections main "Main ATT" subpop "Subpopulation ATTs"
collect title "My title"

which produces:

. collect preview

My title
-----------------------------------------------
                         |         ATT [95% CI]
-------------------------+---------------------
Main ATT                 |                     
  Intervention # Endline |  0.609 [0.152,1.067]
Subpopulation ATTs       |                     
  Age (categories)       |                     
    15-17                |  0.589 [0.090,1.088]
    18-19                |  0.619 [0.164,1.074]
  Wealth quintile        |                     
    Very poor            |  0.678 [0.231,1.124]
    Poor                 | 0.560 [-0.110,1.230]
    Averga               |  0.748 [0.401,1.095]
    Rich                 | 0.495 [-0.069,1.059]
    Very rich            |  0.574 [0.049,1.100]
-----------------------------------------------
* p<.05

[Kehinde Atoloye]

You can add this code at the end of yours:

collect set main
collect addtags sections[main]
collect set q104cat
collect addtags sections[subpop]
collect set wealth_quintile
collect addtags sections[subpop]


collect combine comb = main q104cat wealth_quintile
collect layout (sections # (colname[1.exposure#1.period] q104cat wealth_quintile)) (result[bci])
collect label dim wealth_quintile "Wealth quintile"
collect label dim q104cat "Age (categories)"
collect label levels sections main "Main ATT" subpop "Subpopulation ATTs"
collect title "My title"

which produces:

. collect preview

My title
-----------------------------------------------
| ATT [95% CI]
-------------------------+---------------------
Main ATT |
Intervention # Endline | 0.609 [0.152,1.067]
Subpopulation ATTs |
Age (categories) |
15-17 | 0.589 [0.090,1.088]
18-19 | 0.619 [0.164,1.074]
Wealth quintile |
Very poor | 0.678 [0.231,1.124]
Poor | 0.560 [-0.110,1.230]
Averga | 0.748 [0.401,1.095]
Rich | 0.495 [-0.069,1.059]
Very rich | 0.574 [0.049,1.100]
-----------------------------------------------
* p<.05

Thank you for the assistance. I am just wondering why the stars are not showing when I applied to the original dataset where I expect stars to show. Below is an output of the results from the original data set.

-----------------------------------------------
                         |         ATT [95% CI]
-------------------------+---------------------
Main ATT                 |                     
  Intervention # Endline | 0.455 [0.305, 0.605]
Subpopulation ATTs       |                     
  Age(categories)        |                     
    15-17                | 0.460 [0.307, 0.614]
    18-19                | 0.453 [0.303, 0.603]
  Wealth quintile        |                     
    Very poor            | 0.461 [0.309, 0.613]
    Poor                 | 0.462 [0.311, 0.613]
    Averga               | 0.464 [0.313, 0.615]
    Rich                 | 0.453 [0.301, 0.605]
    Very rich            | 0.438 [0.290, 0.587]
-----------------------------------------------
* p<.05

[Kehinde Atoloye]

You can add this code at the end of yours:

Never mind with the stars problem. I have resolved it. I was calling _r_b instead of _r_p. Thanks.

collect stars _r_p 0.05 "*", attach(_r_b) shownote

My title
------------------------------------------------
                         |          ATT [95% CI]
-------------------------+----------------------
Main ATT                 |                      
  Intervention # Endline | 0.455* [0.305, 0.605]
Subpopulation ATTs       |                      
  Age(categories)        |                      
    15-17                | 0.460* [0.307, 0.614]
    18-19                | 0.453* [0.303, 0.603]
  Wealth quintile        |                      
    Very poor            | 0.461* [0.309, 0.613]
    Poor                 | 0.462* [0.311, 0.613]
    Averga               | 0.464* [0.313, 0.615]
    Rich                 | 0.453* [0.301, 0.605]
    Very rich            | 0.438* [0.290, 0.587]
------------------------------------------------
* p<.05