Announcement

Collapse
No announcement yet.
X
Collapse
  • Page of 2
  • Filter
  • Time
  • Show
Clear All
new posts
  • #1

    Using coefplot to plot odds ratios with p-values label after mi estimate, cmdok : melogit

    I would like to plot odds ratios of mixed-effects logistic models of the association of four sexual risk-taking behaviours (Sex after substance use "SexDrunkDrugs", infrequent condom use "SexLastUnsafeConsist", multiple sexual partnership "SexPartnerFreq2More", and inequitable sexual partnership "InequitSex") with mobile phone use for health content "XC" and mobile phone use for social media "XM" by gender (boys and girls). I also want to show the p-values labels on the graphs.

    I used and adapted Ben Jann coefplot at http://repec.sowi.unibe.ch/stata/coe...g-started.html for estimates from multiply imputed data. Here is the code:

    Code:
    local y     SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex
local x1     XC XM
local xsex    Rural AgeC HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j

foreach ys of local y {
    forvalues f = 0/1 {
        mi estimate, saving(`ys'`f', replace) cmdok : melogit `ys' `x1' `xsex' if Sex == `f' || ID:, or cov(un) 
        estimate store `ys'`f'
    }
}

coefplot    (SexDrunkDrugs0, label("Boy") msymbol(S) mcolor(black) mfcolor(white))     (SexDrunkDrugs1, label("Girl") msymbol(O) mcolor(black) mfcolor(black))    , bylabel("Sex after substance use") mlabel("{it:p} = " + string(@pval,"%9.3f")) mlabpos(1)                                            ///
      ||    (SexLastUnsafeConsist0)            (SexLastUnsafeConsist1)            , bylabel("Infrequent condom use")                 ///
      ||    (SexPartnerFreq2More0)            (SexPartnerFreq2More1)            , bylabel("Multiple sexual partnership")        ///
      ||     (InequitSex0)                    (InequitSex1)                    , bylabel("Inequitable sexual partnership")        ///
      ||      , eform keep(XC XM) ciopts(lcolor(gs8) lwidth(thick)) xtitle("Adjusted odds ratios (ORs)") xlab(0(1)4,format(%3.0fc)) xline(1, lcolor(gs10)) ylab(1 `" "Health" "content" "use" "(Ref. no access)" "' 2 `" "Social" "media" "use" "(Ref. no access)" "') byopts(xrescale) subtitle(, fcolor(none) lstyle(none)) legend(order(2 "ORs for boys" 4 "ORs for girls" 3 "95% CI") row(1) ring(0) pos(10) region(lstyle(none)))
    My data looks like
    Code:
    * Example generated by -dataex-. For more info, type help dataex
clear
input byte(SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex) float(XC XM) byte(Sex Rural) double AgeC byte(HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j) long ID int _mi_id byte _mi_miss int _mi_m
0 0 0 0 0 0 0 1  -.8304964539007091 1 1 1 0 1 0  10   1 0 0
0 1 0 . 1 1 0 1  -.8304964539007091 1 1 1 0 1 1  10   2 1 0
0 1 1 1 0 0 1 0  -.8304964539007091 0 1 1 0 1 0  20   3 0 0
0 1 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 1  20   4 0 0
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 0  50   5 0 0
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 1  50   6 0 0
. 0 0 0 1 1 1 0  .16950354609929086 0 0 1 0 1 0  60   7 1 0
0 1 0 0 0 1 1 0  .16950354609929086 0 0 1 0 1 1  60   8 0 0
0 0 0 0 0 0 0 1  -.8304964539007091 0 1 0 0 1 0  70   9 0 0
0 0 0 0 1 1 0 1  -.8304964539007091 0 1 0 0 1 1  70  10 0 0
0 0 0 0 0 0 1 0  -.8304964539007091 1 0 1 0 1 0  90  11 0 0
0 0 0 0 0 0 1 0  -.8304964539007091 1 0 1 0 1 1  90  12 0 0
0 0 0 0 0 0 0 1 -1.8304964539007091 0 0 1 0 1 0 100  13 0 0
0 0 0 0 0 0 0 1 -1.8304964539007091 0 0 1 0 1 1 100  14 0 0
0 0 0 0 0 1 0 0  -2.830496453900709 0 0 1 0 1 0 140  15 0 0
0 0 0 0 0 0 0 0  -2.830496453900709 0 0 1 0 1 1 140  16 0 0
0 0 0 0 0 0 1 0  -2.830496453900709 0 0 0 0 1 0 150  17 0 0
0 0 0 0 0 1 1 0  -2.830496453900709 0 0 0 0 1 1 150  18 0 0
. 0 0 0 1 1 0 0  -.8304964539007091 0 0 1 1 1 0 170  19 1 0
1 1 1 . 0 1 0 0  -.8304964539007091 0 0 1 1 1 1 170  20 1 0
0 0 0 0 0 1 1 1  -.8304964539007091 0 0 1 0 1 0 200  21 0 0
0 0 0 0 0 1 1 1  -.8304964539007091 0 0 1 0 1 1 200  22 0 0
. 0 0 0 0 0 0 0  -.8304964539007091 0 0 0 0 1 0 210  23 1 0
0 0 0 0 0 0 0 0  -.8304964539007091 0 0 0 0 1 1 210  24 0 0
0 0 0 0 0 1 1 1 -1.8304964539007091 0 0 1 0 1 0 230  25 0 0
0 0 0 0 0 1 1 1 -1.8304964539007091 0 0 1 0 1 1 230  26 0 0
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 0 240  27 0 0
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 1 240  28 0 0
0 0 0 0 0 1 1 1  .16950354609929086 0 1 0 0 1 0 280  29 0 0
0 0 0 0 0 0 1 1  .16950354609929086 0 1 0 0 1 1 280  30 0 0
0 0 0 0 0 0 0 0 -1.8304964539007091 0 1 1 0 1 0 290  31 0 0
0 0 0 0 0 0 0 0 -1.8304964539007091 0 1 1 0 1 1 290  32 0 0
0 0 1 0 0 0 0 0   6.169503546099291 0 0 1 1 1 0 300  33 0 0
0 0 1 0 0 0 0 0   6.169503546099291 0 0 1 1 1 1 300  34 0 0
0 0 0 0 0 0 1 0   3.169503546099291 0 0 1 1 1 0 310  35 0 0
0 0 0 0 0 0 1 0   3.169503546099291 0 0 1 1 1 1 310  36 0 0
0 0 0 0 0 0 0 0  -.8304964539007091 0 1 0 0 1 0 340  37 0 0
1 1 0 . 0 0 0 0  -.8304964539007091 0 1 0 0 1 1 340  38 1 0
. 0 0 0 0 1 1 0   3.169503546099291 0 0 1 0 1 0 360  39 1 0
0 0 0 0 0 1 1 0   3.169503546099291 0 0 1 0 1 1 360  40 0 0
0 0 0 0 0 1 1 0  .16950354609929086 0 0 1 0 1 0 390  41 0 0
0 0 0 0 0 0 1 0  .16950354609929086 0 0 1 0 1 1 390  42 0 0
0 0 0 0 0 0 1 0 -1.8304964539007091 0 0 1 0 1 0 410  43 0 0
0 0 0 0 0 1 1 0 -1.8304964539007091 0 0 1 0 1 1 410  44 0 0
0 0 0 0 0 0 1 1  -3.830496453900709 0 0 0 0 1 0 440  45 0 0
0 0 0 0 0 1 1 1  -3.830496453900709 0 0 0 0 1 1 440  46 0 0
1 0 1 1 0 1 0 0  .16950354609929086 0 1 1 0 1 0 450  47 0 0
. 1 0 . 0 1 0 0  .16950354609929086 0 1 1 0 1 1 450  48 1 0
0 0 0 0 0 1 1 1  -.8304964539007091 0 0 1 0 1 0 470  49 0 0
0 0 0 0 0 0 1 1  -.8304964539007091 0 0 1 0 1 1 470  50 0 0
0 0 0 0 0 0 0 1  -3.830496453900709 1 1 1 0 1 0 480  51 0 0
0 0 0 0 0 1 0 1  -3.830496453900709 1 1 1 0 1 1 480  52 0 0
0 0 0 0 0 1 1 0   3.169503546099291 0 0 0 0 1 0 490  53 0 0
0 1 0 0 0 1 1 0   3.169503546099291 0 0 0 0 1 1 490  54 0 0
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 0 1 1 0 510  55 0 0
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 0 1 1 1 510  56 0 0
0 0 1 0 0 1 1 0  1.1695035460992909 0 0 0 1 1 0 530  57 0 0
0 0 1 1 0 1 1 0  1.1695035460992909 0 0 0 1 1 1 530  58 0 0
0 0 1 0 0 0 1 0  1.1695035460992909 0 0 1 0 1 0 540  59 0 0
0 1 0 1 0 1 1 0  1.1695035460992909 0 0 1 0 1 1 540  60 0 0
0 0 1 0 0 0 1 1   3.169503546099291 0 1 1 1 1 0 580  61 0 0
0 1 1 1 0 1 1 1   3.169503546099291 0 1 1 1 1 1 580  62 0 0
0 0 1 0 0 1 1 1   2.169503546099291 0 0 1 1 1 0 600  63 0 0
0 0 0 0 0 1 1 1   2.169503546099291 0 0 1 1 1 1 600  64 0 0
0 0 0 0 0 0 1 0  -.8304964539007091 0 0 1 0 1 0 620  65 0 0
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 1 0 1 1 620  66 0 0
0 1 1 0 0 0 1 0  .16950354609929086 0 0 0 1 1 0 650  67 0 0
0 1 0 0 0 1 1 0  .16950354609929086 0 0 0 1 1 1 650  68 0 0
1 1 1 0 0 0 1 0  -.8304964539007091 0 0 0 0 1 0 680  69 0 0
0 0 0 1 0 0 1 0  -.8304964539007091 0 0 0 0 1 1 680  70 0 0
0 0 0 0 0 0 0 0  -3.830496453900709 0 1 1 0 1 0 690  71 0 0
. 0 0 . 0 0 0 0  -3.830496453900709 0 1 1 0 1 1 690  72 1 0
0 0 0 0 0 0 1 0  -2.830496453900709 0 1 0 0 1 0 700  73 0 0
0 0 0 0 0 0 1 0  -2.830496453900709 0 1 0 0 1 1 700  74 0 0
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 0 0 1 0 710  75 0 0
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 0 0 1 1 710  76 0 0
0 0 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 0 750  77 0 0
0 0 0 0 0 1 1 0  -.8304964539007091 0 1 1 0 1 1 750  78 0 0
0 0 0 0 1 1 1 1   2.169503546099291 0 0 1 1 1 0 780  79 0 0
0 1 0 0 0 1 1 1   2.169503546099291 0 0 1 1 1 1 780  80 0 0
. 0 0 0 0 1 0 0  1.1695035460992909 0 0 0 1 1 0 790  81 1 0
0 1 0 0 0 1 0 0  1.1695035460992909 0 0 0 1 1 1 790  82 0 0
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 1 0 1 0 800  83 0 0
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 1 0 1 1 800  84 0 0
. 0 0 0 0 0 1 0  1.1695035460992909 0 0 0 1 1 0 810  85 1 0
0 0 0 0 0 1 1 0  1.1695035460992909 0 0 0 1 1 1 810  86 0 0
. 0 0 0 0 1 1 1  .16950354609929086 0 1 0 0 1 0 830  87 1 0
0 1 0 0 1 1 1 1  .16950354609929086 0 1 0 0 1 1 830  88 0 0
0 0 0 0 0 0 1 0 -1.8304964539007091 0 1 0 0 1 0 850  89 0 0
0 1 0 . 1 1 1 0 -1.8304964539007091 0 1 0 0 1 1 850  90 1 0
0 0 0 0 0 0 1 1  -2.830496453900709 0 1 0 0 1 0 870  91 0 0
0 0 0 0 0 0 1 1  -2.830496453900709 0 1 0 0 1 1 870  92 0 0
0 0 0 0 1 1 1 0  1.1695035460992909 0 0 0 0 1 0 930  93 0 0
0 0 0 0 0 0 1 0  1.1695035460992909 0 0 0 0 1 1 930  94 0 0
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 1 0 1 0 940  95 0 0
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 1 0 1 1 940  96 0 0
. 0 0 0 0 0 1 0  .16950354609929086 1 0 0 1 1 0 950  97 1 0
0 1 0 0 0 0 1 0  .16950354609929086 1 0 0 1 1 1 950  98 0 0
0 0 0 0 0 0 1 1  -3.830496453900709 0 0 1 0 1 0 970  99 0 0
0 0 0 0 0 0 1 1  -3.830496453900709 0 0 1 0 1 1 970 100 0 0
end
label values SexLastUnsafeConsist yes_no
label values SexPartnerFreq2More yes_no
label def yes_no 0 "0. No", modify
label def yes_no 1 "1. Yes", modify
label values Sex Sex
label def Sex 0 "Male", modify
label def Sex 1 "Female", modify
label values Rural UrbRur
label def UrbRur 0 "0. Urban", modify
label def UrbRur 1 "1. Rural", modify
label values ARTever binary
label values HomeTypeInformal binary
label values Relationship binary
label values SchEnrol binary
label def binary 0 "0. No", modify
label def binary 1 "1. Yes", modify
label values NecessitiesAll_r binary_r
label def binary_r 0 "0. Yes", modify
label def binary_r 1 "1. No", modify
    I got a figure like the one I attached to this post Graph.gph .

    Please, is there any reason why the p-value for health content use for infrequent condom use is missing.

    I tried the same coding for non-imputed dataset and everything works.

    I would appreciate any help from you.

    Best,
    Boladé
    Tags: cmdok : melogit, coefplot, graph, graphics, mi estimate
    • 1 like
  • #2
    Perhaps you need the -post- option of mi estimate. It's not possible to replicate the issue without the imputation codes.

    Code:
    mi estimate, post saving(`ys'`f', replace) cmdok : melogit ...
    • 2 likes

    Comment

    • #3
      Thank you Andrew. Sorry I thought the imputed data would be enough. The imputation code is
      Code:
      * Missingness
forvalues i = 2/3 {
    mdesc `SRH_4' Use Rural AgeC Sex ARTever Relationship SchEnrol NecessitiesAll_r HomeTypeInformal if Wave == `i'
    misstable patterns `SRH_4' Use Rural AgeC Sex ARTever Relationship SchEnrol NecessitiesAll_r HomeTypeInformal if Wave == `i'
}

* The amount of missing data for the key variables at T2 and T3 (for included participants) were, respectively: 13.5% and 6.5% for sex after substance use; 3.8% and 8.6% for inequitable sexual partnership. There is no missing data on mobile phone access and use (neither at T2, nor at T3). The amount of missing data for covariates (for included participants) at baseline was 0.1% for household poverty informal housing and 0% for sex, age, rural residence, relationship status, HIV status and school enrolment
/*
Missing-value patterns
     (1 means complete)

* at Wave 2
              |   Pattern
    Percent   |  1  2  3  4
  ------------+-------------
       85%    |  1  1  1  1
              |
       11     |  1  1  1  0
        3     |  1  1  0  0
        1     |  1  1  0  1
       <1     |  0  1  1  1
       <1     |  1  0  1  1
  ------------+-------------
      100%    |

 Variables are  (1) HomeTypeInformal  (2) NecessitiesAll_r  (3) InequitSex  (4) SexDrunkDrugs

Interpretation: At T2, the dataset contains missing values four variables including household poverty, informal housing, inequitable sexual partnership, and sex after substance use. 85% of data are complete while 15% of the observations contain missing values.

* at Wave 3
              |   Pattern
    Percent   |  1  2  3  4
  ------------+-------------
       91%    |  1  1  1  1
              |
        7     |  1  1  0  0
        2     |  1  1  1  0
       <1     |  0  1  1  1
       <1     |  1  0  1  1
  ------------+-------------
      100%    |

Variables are  (1) HomeTypeInformal  (2) NecessitiesAll_r  (3) SexDrunkDrugs  (4) InequitSex

Interpretation: At T3, the dataset contains missing values four variables including household poverty, informal housing, sex after substance use, and inequitable sexual partnership. 91% of data are complete while 9% of the observations contain missing values.
*/

* Now we can determine whether the data are MCAR using the mcartest command.

forvalues i = 2/3 {
    mcartest SexDrunkDrugs InequitSex = i.HomeTypeInformal i.NecessitiesAll_r i.XC i.XM i.Rural AgeC i.Sex i.ARTever i.SchEnrol i.Relationship if Wave == `i', emoutput nolog
}

* Little's χ² test indicated that data were not missing completely at random (appendix x2)

* Missing data are not MCAR

mi set mlong    // mi set data for imputation

* Reshape data to wide (it makes it easly to use variables from T2 as predictors of missing values)

* Firt drop variables not needed in analysis
drop SexIntercourseEverProxy q1* XCXM SRH_count HIVRiskAny HIVRisk2Form HIVRiskAll Wave wgt touse-ibInequitSex

mi reshape wide SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)

* Identifying which variables in the imputation model have missing information.
mi register imputed NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0 SexDrunkDrugs1 InequitSex0 InequitSex1

*** Mutiple imputation
*MI using chained equations/MICE (also known as the fully conditional specification or sequential generalized regression) for binary outcomes.  In simulation studies (Lee & Carlin, 2010; Van Buuren, 2007), MICE has been show to produce estimates that are comparable to MVN method.

* Clear any xtset that is ongoing
mi xtset, clear
mi /*mimpt*/ impute chained     (logit) NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0     ///
                    SexDrunkDrugs1 InequitSex0 InequitSex1 = Access0 Access1     ///
                    XC0 XM0 XC1 XM1 Sex ARTever AgeC Rural Relationship         ///
                    SchEnrol, add(10) rseed(20210101) /*skipnonconvergence(5)*/ savetrace(tracedata, replace)

* Bring data back to long format
mi reshape long SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)

* We can check the data and see if there is no problems in our imputation
summ SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex HomeTypeInformal NecessitiesAll_r
      And the data before imputation is

      Code:
      * Example generated by -dataex-. For more info, type help dataex
clear
input byte(SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More) float(InequitSex XC XM) byte(Sex Rural) double AgeC byte(HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j) long ID
0 0 0 0 0 0 0 1  -.8304964539007091 1 1 1 0 1 0  10
0 1 0 . 1 1 0 1  -.8304964539007091 1 1 1 0 1 1  10
0 1 1 1 0 0 1 0  -.8304964539007091 0 1 1 0 1 0  20
0 1 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 1  20
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 0  50
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 1  50
. 0 0 0 1 1 1 0  .16950354609929086 0 0 1 0 1 0  60
0 1 0 0 0 1 1 0  .16950354609929086 0 0 1 0 1 1  60
0 0 0 0 0 0 0 1  -.8304964539007091 0 1 0 0 1 0  70
0 0 0 0 1 1 0 1  -.8304964539007091 0 1 0 0 1 1  70
0 0 0 0 0 0 1 0  -.8304964539007091 1 0 1 0 1 0  90
0 0 0 0 0 0 1 0  -.8304964539007091 1 0 1 0 1 1  90
0 0 0 0 0 0 0 1 -1.8304964539007091 0 0 1 0 1 0 100
0 0 0 0 0 0 0 1 -1.8304964539007091 0 0 1 0 1 1 100
0 0 0 0 0 1 0 0  -2.830496453900709 0 0 1 0 1 0 140
0 0 0 0 0 0 0 0  -2.830496453900709 0 0 1 0 1 1 140
0 0 0 0 0 0 1 0  -2.830496453900709 0 0 0 0 1 0 150
0 0 0 0 0 1 1 0  -2.830496453900709 0 0 0 0 1 1 150
. 0 0 0 1 1 0 0  -.8304964539007091 0 0 1 1 1 0 170
1 1 1 . 0 1 0 0  -.8304964539007091 0 0 1 1 1 1 170
0 0 0 0 0 1 1 1  -.8304964539007091 0 0 1 0 1 0 200
0 0 0 0 0 1 1 1  -.8304964539007091 0 0 1 0 1 1 200
. 0 0 0 0 0 0 0  -.8304964539007091 0 0 0 0 1 0 210
0 0 0 0 0 0 0 0  -.8304964539007091 0 0 0 0 1 1 210
0 0 0 0 0 1 1 1 -1.8304964539007091 0 0 1 0 1 0 230
0 0 0 0 0 1 1 1 -1.8304964539007091 0 0 1 0 1 1 230
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 0 240
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 1 240
0 0 0 0 0 1 1 1  .16950354609929086 0 1 0 0 1 0 280
0 0 0 0 0 0 1 1  .16950354609929086 0 1 0 0 1 1 280
0 0 0 0 0 0 0 0 -1.8304964539007091 0 1 1 0 1 0 290
0 0 0 0 0 0 0 0 -1.8304964539007091 0 1 1 0 1 1 290
0 0 1 0 0 0 0 0   6.169503546099291 0 0 1 1 1 0 300
0 0 1 0 0 0 0 0   6.169503546099291 0 0 1 1 1 1 300
0 0 0 0 0 0 1 0   3.169503546099291 0 0 1 1 1 0 310
0 0 0 0 0 0 1 0   3.169503546099291 0 0 1 1 1 1 310
0 0 0 0 0 0 0 0  -.8304964539007091 0 1 0 0 1 0 340
1 1 0 . 0 0 0 0  -.8304964539007091 0 1 0 0 1 1 340
. 0 0 0 0 1 1 0   3.169503546099291 0 0 1 0 1 0 360
0 0 0 0 0 1 1 0   3.169503546099291 0 0 1 0 1 1 360
0 0 0 0 0 1 1 0  .16950354609929086 0 0 1 0 1 0 390
0 0 0 0 0 0 1 0  .16950354609929086 0 0 1 0 1 1 390
0 0 0 0 0 0 1 0 -1.8304964539007091 0 0 1 0 1 0 410
0 0 0 0 0 1 1 0 -1.8304964539007091 0 0 1 0 1 1 410
0 0 0 0 0 0 1 1  -3.830496453900709 0 0 0 0 1 0 440
0 0 0 0 0 1 1 1  -3.830496453900709 0 0 0 0 1 1 440
1 0 1 1 0 1 0 0  .16950354609929086 0 1 1 0 1 0 450
. 1 0 . 0 1 0 0  .16950354609929086 0 1 1 0 1 1 450
0 0 0 0 0 1 1 1  -.8304964539007091 0 0 1 0 1 0 470
0 0 0 0 0 0 1 1  -.8304964539007091 0 0 1 0 1 1 470
0 0 0 0 0 0 0 1  -3.830496453900709 1 1 1 0 1 0 480
0 0 0 0 0 1 0 1  -3.830496453900709 1 1 1 0 1 1 480
0 0 0 0 0 1 1 0   3.169503546099291 0 0 0 0 1 0 490
0 1 0 0 0 1 1 0   3.169503546099291 0 0 0 0 1 1 490
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 0 1 1 0 510
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 0 1 1 1 510
0 0 1 0 0 1 1 0  1.1695035460992909 0 0 0 1 1 0 530
0 0 1 1 0 1 1 0  1.1695035460992909 0 0 0 1 1 1 530
0 0 1 0 0 0 1 0  1.1695035460992909 0 0 1 0 1 0 540
0 1 0 1 0 1 1 0  1.1695035460992909 0 0 1 0 1 1 540
0 0 1 0 0 0 1 1   3.169503546099291 0 1 1 1 1 0 580
0 1 1 1 0 1 1 1   3.169503546099291 0 1 1 1 1 1 580
0 0 1 0 0 1 1 1   2.169503546099291 0 0 1 1 1 0 600
0 0 0 0 0 1 1 1   2.169503546099291 0 0 1 1 1 1 600
0 0 0 0 0 0 1 0  -.8304964539007091 0 0 1 0 1 0 620
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 1 0 1 1 620
0 1 1 0 0 0 1 0  .16950354609929086 0 0 0 1 1 0 650
0 1 0 0 0 1 1 0  .16950354609929086 0 0 0 1 1 1 650
1 1 1 0 0 0 1 0  -.8304964539007091 0 0 0 0 1 0 680
0 0 0 1 0 0 1 0  -.8304964539007091 0 0 0 0 1 1 680
0 0 0 0 0 0 0 0  -3.830496453900709 0 1 1 0 1 0 690
. 0 0 . 0 0 0 0  -3.830496453900709 0 1 1 0 1 1 690
0 0 0 0 0 0 1 0  -2.830496453900709 0 1 0 0 1 0 700
0 0 0 0 0 0 1 0  -2.830496453900709 0 1 0 0 1 1 700
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 0 0 1 0 710
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 0 0 1 1 710
0 0 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 0 750
0 0 0 0 0 1 1 0  -.8304964539007091 0 1 1 0 1 1 750
0 0 0 0 1 1 1 1   2.169503546099291 0 0 1 1 1 0 780
0 1 0 0 0 1 1 1   2.169503546099291 0 0 1 1 1 1 780
. 0 0 0 0 1 0 0  1.1695035460992909 0 0 0 1 1 0 790
0 1 0 0 0 1 0 0  1.1695035460992909 0 0 0 1 1 1 790
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 1 0 1 0 800
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 1 0 1 1 800
. 0 0 0 0 0 1 0  1.1695035460992909 0 0 0 1 1 0 810
0 0 0 0 0 1 1 0  1.1695035460992909 0 0 0 1 1 1 810
. 0 0 0 0 1 1 1  .16950354609929086 0 1 0 0 1 0 830
0 1 0 0 1 1 1 1  .16950354609929086 0 1 0 0 1 1 830
0 0 0 0 0 0 1 0 -1.8304964539007091 0 1 0 0 1 0 850
0 1 0 . 1 1 1 0 -1.8304964539007091 0 1 0 0 1 1 850
0 0 0 0 0 0 1 1  -2.830496453900709 0 1 0 0 1 0 870
0 0 0 0 0 0 1 1  -2.830496453900709 0 1 0 0 1 1 870
0 0 0 0 1 1 1 0  1.1695035460992909 0 0 0 0 1 0 930
0 0 0 0 0 0 1 0  1.1695035460992909 0 0 0 0 1 1 930
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 1 0 1 0 940
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 1 0 1 1 940
. 0 0 0 0 0 1 0  .16950354609929086 1 0 0 1 1 0 950
0 1 0 0 0 0 1 0  .16950354609929086 1 0 0 1 1 1 950
0 0 0 0 0 0 1 1  -3.830496453900709 0 0 1 0 1 0 970
0 0 0 0 0 0 1 1  -3.830496453900709 0 0 1 0 1 1 970
end
label values SexLastUnsafeConsist yes_no
label values SexPartnerFreq2More yes_no
label def yes_no 0 "0. No", modify
label def yes_no 1 "1. Yes", modify
label values Sex Sex
label def Sex 0 "Male", modify
label def Sex 1 "Female", modify
label values Rural UrbRur
label def UrbRur 0 "0. Urban", modify
label def UrbRur 1 "1. Rural", modify
label values ARTever binary
label values HomeTypeInformal binary
label values Relationship binary
label values SchEnrol binary
label def binary 0 "0. No", modify
label def binary 1 "1. Yes", modify
label values NecessitiesAll_r binary_r
label def binary_r 0 "0. Yes", modify
label def binary_r 1 "1. No", modify
      I hope this helps.

      Best,
      Boladé

      Comment

      • #4
        Hi again Andrew, I try to add the post option and still get the same output.

        Comment

        • #5
          When I run the following

          Code:
          * Example generated by -dataex-. For more info, type help dataex
clear
input byte(SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More) float(InequitSex XC XM) byte(Sex Rural) double AgeC byte(HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j) long ID
0 0 0 0 0 0 0 1  -.8304964539007091 1 1 1 0 1 0  10
0 1 0 . 1 1 0 1  -.8304964539007091 1 1 1 0 1 1  10
0 1 1 1 0 0 1 0  -.8304964539007091 0 1 1 0 1 0  20
0 1 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 1  20
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 0  50
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 1  50
. 0 0 0 1 1 1 0  .16950354609929086 0 0 1 0 1 0  60
0 1 0 0 0 1 1 0  .16950354609929086 0 0 1 0 1 1  60
0 0 0 0 0 0 0 1  -.8304964539007091 0 1 0 0 1 0  70
0 0 0 0 1 1 0 1  -.8304964539007091 0 1 0 0 1 1  70
0 0 0 0 0 0 1 0  -.8304964539007091 1 0 1 0 1 0  90
0 0 0 0 0 0 1 0  -.8304964539007091 1 0 1 0 1 1  90
0 0 0 0 0 0 0 1 -1.8304964539007091 0 0 1 0 1 0 100
0 0 0 0 0 0 0 1 -1.8304964539007091 0 0 1 0 1 1 100
0 0 0 0 0 1 0 0  -2.830496453900709 0 0 1 0 1 0 140
0 0 0 0 0 0 0 0  -2.830496453900709 0 0 1 0 1 1 140
0 0 0 0 0 0 1 0  -2.830496453900709 0 0 0 0 1 0 150
0 0 0 0 0 1 1 0  -2.830496453900709 0 0 0 0 1 1 150
. 0 0 0 1 1 0 0  -.8304964539007091 0 0 1 1 1 0 170
1 1 1 . 0 1 0 0  -.8304964539007091 0 0 1 1 1 1 170
0 0 0 0 0 1 1 1  -.8304964539007091 0 0 1 0 1 0 200
0 0 0 0 0 1 1 1  -.8304964539007091 0 0 1 0 1 1 200
. 0 0 0 0 0 0 0  -.8304964539007091 0 0 0 0 1 0 210
0 0 0 0 0 0 0 0  -.8304964539007091 0 0 0 0 1 1 210
0 0 0 0 0 1 1 1 -1.8304964539007091 0 0 1 0 1 0 230
0 0 0 0 0 1 1 1 -1.8304964539007091 0 0 1 0 1 1 230
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 0 240
0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 1 240
0 0 0 0 0 1 1 1  .16950354609929086 0 1 0 0 1 0 280
0 0 0 0 0 0 1 1  .16950354609929086 0 1 0 0 1 1 280
0 0 0 0 0 0 0 0 -1.8304964539007091 0 1 1 0 1 0 290
0 0 0 0 0 0 0 0 -1.8304964539007091 0 1 1 0 1 1 290
0 0 1 0 0 0 0 0   6.169503546099291 0 0 1 1 1 0 300
0 0 1 0 0 0 0 0   6.169503546099291 0 0 1 1 1 1 300
0 0 0 0 0 0 1 0   3.169503546099291 0 0 1 1 1 0 310
0 0 0 0 0 0 1 0   3.169503546099291 0 0 1 1 1 1 310
0 0 0 0 0 0 0 0  -.8304964539007091 0 1 0 0 1 0 340
1 1 0 . 0 0 0 0  -.8304964539007091 0 1 0 0 1 1 340
. 0 0 0 0 1 1 0   3.169503546099291 0 0 1 0 1 0 360
0 0 0 0 0 1 1 0   3.169503546099291 0 0 1 0 1 1 360
0 0 0 0 0 1 1 0  .16950354609929086 0 0 1 0 1 0 390
0 0 0 0 0 0 1 0  .16950354609929086 0 0 1 0 1 1 390
0 0 0 0 0 0 1 0 -1.8304964539007091 0 0 1 0 1 0 410
0 0 0 0 0 1 1 0 -1.8304964539007091 0 0 1 0 1 1 410
0 0 0 0 0 0 1 1  -3.830496453900709 0 0 0 0 1 0 440
0 0 0 0 0 1 1 1  -3.830496453900709 0 0 0 0 1 1 440
1 0 1 1 0 1 0 0  .16950354609929086 0 1 1 0 1 0 450
. 1 0 . 0 1 0 0  .16950354609929086 0 1 1 0 1 1 450
0 0 0 0 0 1 1 1  -.8304964539007091 0 0 1 0 1 0 470
0 0 0 0 0 0 1 1  -.8304964539007091 0 0 1 0 1 1 470
0 0 0 0 0 0 0 1  -3.830496453900709 1 1 1 0 1 0 480
0 0 0 0 0 1 0 1  -3.830496453900709 1 1 1 0 1 1 480
0 0 0 0 0 1 1 0   3.169503546099291 0 0 0 0 1 0 490
0 1 0 0 0 1 1 0   3.169503546099291 0 0 0 0 1 1 490
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 0 1 1 0 510
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 0 1 1 1 510
0 0 1 0 0 1 1 0  1.1695035460992909 0 0 0 1 1 0 530
0 0 1 1 0 1 1 0  1.1695035460992909 0 0 0 1 1 1 530
0 0 1 0 0 0 1 0  1.1695035460992909 0 0 1 0 1 0 540
0 1 0 1 0 1 1 0  1.1695035460992909 0 0 1 0 1 1 540
0 0 1 0 0 0 1 1   3.169503546099291 0 1 1 1 1 0 580
0 1 1 1 0 1 1 1   3.169503546099291 0 1 1 1 1 1 580
0 0 1 0 0 1 1 1   2.169503546099291 0 0 1 1 1 0 600
0 0 0 0 0 1 1 1   2.169503546099291 0 0 1 1 1 1 600
0 0 0 0 0 0 1 0  -.8304964539007091 0 0 1 0 1 0 620
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 1 0 1 1 620
0 1 1 0 0 0 1 0  .16950354609929086 0 0 0 1 1 0 650
0 1 0 0 0 1 1 0  .16950354609929086 0 0 0 1 1 1 650
1 1 1 0 0 0 1 0  -.8304964539007091 0 0 0 0 1 0 680
0 0 0 1 0 0 1 0  -.8304964539007091 0 0 0 0 1 1 680
0 0 0 0 0 0 0 0  -3.830496453900709 0 1 1 0 1 0 690
. 0 0 . 0 0 0 0  -3.830496453900709 0 1 1 0 1 1 690
0 0 0 0 0 0 1 0  -2.830496453900709 0 1 0 0 1 0 700
0 0 0 0 0 0 1 0  -2.830496453900709 0 1 0 0 1 1 700
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 0 0 1 0 710
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 0 0 1 1 710
0 0 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 0 750
0 0 0 0 0 1 1 0  -.8304964539007091 0 1 1 0 1 1 750
0 0 0 0 1 1 1 1   2.169503546099291 0 0 1 1 1 0 780
0 1 0 0 0 1 1 1   2.169503546099291 0 0 1 1 1 1 780
. 0 0 0 0 1 0 0  1.1695035460992909 0 0 0 1 1 0 790
0 1 0 0 0 1 0 0  1.1695035460992909 0 0 0 1 1 1 790
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 1 0 1 0 800
0 0 0 0 0 1 1 0  -.8304964539007091 0 0 1 0 1 1 800
. 0 0 0 0 0 1 0  1.1695035460992909 0 0 0 1 1 0 810
0 0 0 0 0 1 1 0  1.1695035460992909 0 0 0 1 1 1 810
. 0 0 0 0 1 1 1  .16950354609929086 0 1 0 0 1 0 830
0 1 0 0 1 1 1 1  .16950354609929086 0 1 0 0 1 1 830
0 0 0 0 0 0 1 0 -1.8304964539007091 0 1 0 0 1 0 850
0 1 0 . 1 1 1 0 -1.8304964539007091 0 1 0 0 1 1 850
0 0 0 0 0 0 1 1  -2.830496453900709 0 1 0 0 1 0 870
0 0 0 0 0 0 1 1  -2.830496453900709 0 1 0 0 1 1 870
0 0 0 0 1 1 1 0  1.1695035460992909 0 0 0 0 1 0 930
0 0 0 0 0 0 1 0  1.1695035460992909 0 0 0 0 1 1 930
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 1 0 1 0 940
0 0 0 0 0 0 1 0  -3.830496453900709 0 0 1 0 1 1 940
. 0 0 0 0 0 1 0  .16950354609929086 1 0 0 1 1 0 950
0 1 0 0 0 0 1 0  .16950354609929086 1 0 0 1 1 1 950
0 0 0 0 0 0 1 1  -3.830496453900709 0 0 1 0 1 0 970
0 0 0 0 0 0 1 1  -3.830496453900709 0 0 1 0 1 1 970
end
label values SexLastUnsafeConsist yes_no
label values SexPartnerFreq2More yes_no
label def yes_no 0 "0. No", modify
label def yes_no 1 "1. Yes", modify
label values Sex Sex
label def Sex 0 "Male", modify
label def Sex 1 "Female", modify
label values Rural UrbRur
label def UrbRur 0 "0. Urban", modify
label def UrbRur 1 "1. Rural", modify
label values ARTever binary
label values HomeTypeInformal binary
label values Relationship binary
label values SchEnrol binary
label def binary 0 "0. No", modify
label def binary 1 "1. Yes", modify
label values NecessitiesAll_r binary_r
label def binary_r 0 "0. Yes", modify
label def binary_r 1 "1. No", modify

mi set mlong    // mi set data for imputation

* Reshape data to wide (it makes it easly to use variables from T2 as predictors of missing values)

* Firt drop variables not needed in analysis
drop SexIntercourseEverProxy q1* XCXM SRH_count HIVRiskAny HIVRisk2Form HIVRiskAll Wave wgt touse-ibInequitSex

mi reshape wide SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)

* Identifying which variables in the imputation model have missing information.
mi register imputed NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0 SexDrunkDrugs1 InequitSex0 InequitSex1

*** Mutiple imputation
*MI using chained equations/MICE (also known as the fully conditional specification or sequential generalized regression) for binary outcomes.  In simulation studies (Lee & Carlin, 2010; Van Buuren, 2007), MICE has been show to produce estimates that are comparable to MVN method.

* Clear any xtset that is ongoing
mi xtset, clear
mi /*mimpt*/ impute chained     (logit) NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0     ///
                    SexDrunkDrugs1 InequitSex0 InequitSex1 = Access0 Access1     ///
                    XC0 XM0 XC1 XM1 Sex ARTever AgeC Rural Relationship         ///
                    SchEnrol, add(10) rseed(20210101) /*skipnonconvergence(5)*/ savetrace(tracedata, replace)

* Bring data back to long format
mi reshape long SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)

* We can check the data and see if there is no problems in our imputation
summ SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex HomeTypeInformal NecessitiesAll_r

local y     SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex
local x1     XC XM
local xsex    Rural AgeC HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j

foreach ys of local y {
    forvalues f = 0/1 {
        mi estimate, saving(`ys'`f', replace) cmdok : melogit `ys' `x1' `xsex' if Sex == `f' || ID:, or cov(un) 
        estimate store `ys'`f'
    }
}

coefplot    (SexDrunkDrugs0, label("Boy") msymbol(S) mcolor(black) mfcolor(white))     (SexDrunkDrugs1, label("Girl") msymbol(O) mcolor(black) mfcolor(black))    , bylabel("Sex after substance use") mlabel("{it:p} = " + string(@pval,"%9.3f")) mlabpos(1)                                            ///
      ||    (SexLastUnsafeConsist0)            (SexLastUnsafeConsist1)            , bylabel("Infrequent condom use")                 ///
      ||    (SexPartnerFreq2More0)            (SexPartnerFreq2More1)            , bylabel("Multiple sexual partnership")        ///
      ||     (InequitSex0)                    (InequitSex1)                    , bylabel("Inequitable sexual partnership")        ///
      ||      , eform keep(XC XM) ciopts(lcolor(gs8) lwidth(thick)) xtitle("Adjusted odds ratios (ORs)") xlab(0(1)4,format(%3.0fc)) xline(1, lcolor(gs10)) ylab(1 `" "Health" "content" "use" "(Ref. no access)" "' 2 `" "Social" "media" "use" "(Ref. no access)" "') byopts(xrescale) subtitle(, fcolor(none) lstyle(none)) legend(order(2 "ORs for boys" 4 "ORs for girls" 3 "95% CI") row(1) ring(0) pos(10) region(lstyle(none)))
          I get

          Code:
           
. * Firt drop variables not needed in analysis
. drop SexIntercourseEverProxy q1* XCXM SRH_count HIVRiskAny HIVRisk2Form HIVRiskAll Wave wgt touse-ibInequitSex
variable SexIntercourseEverProxy not found
r(111);

end of do-file
          Make sure that the problem is reproducible, i.e., the code runs without any errors and replicates the issue. Then we can take it from there.

          Comment

          • #6
            Thank you Andrew for your reply. The imputation code is:
            Code:
            * Little's χ² test indicated that data were not missing completely at random (appendix x2)  
  * Missing data are not MCAR
   
  mi set mlong // mi set data for imputation
   
  mi reshape wide SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)
   
  * Identifying which variables in the imputation model have missing information.
  mi register imputed NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0 SexDrunkDrugs1 InequitSex0 InequitSex1
   
  *** Mutiple imputation
  *MI using chained equations/MICE (also known as the fully conditional specification or sequential generalized regression) for binary outcomes.  In simulation studies (Lee & Carlin, 2010; Van Buuren, 2007), MICE has been show to produce estimates that are comparable to MVN method.
   
  * Clear any xtset that is ongoing
  mi xtset, clear
  mi /*mimpt*/ impute chained  (logit) NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0  ///
  SexDrunkDrugs1 InequitSex0 InequitSex1 = Access0 Access1  ///
  XC0 XM0 XC1 XM1 Sex ARTever AgeC Rural Relationship  ///
  SchEnrol, add(10) rseed(20210101) /*skipnonconvergence(5)*/ savetrace(tracedata, replace)
   
  * Bring data back to long format
  mi reshape long SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)
   
  * We can check the data and see if there is no problems in our imputation
 summ SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex HomeTypeInformal NecessitiesAll_r
            on the following data:

            Code:
             * Example generated by -dataex-. For more info, type help dataex
  clear
  input byte(SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More) float(InequitSex Use Access XC XM) byte(Sex Rural) double AgeC byte(HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j) long ID
  0 0 0 0 0 0 0 0 0 1  -.8304964539007091 1 1 1 0 1 0  10
  0 1 0 . 2 1 1 1 0 1  -.8304964539007091 1 1 1 0 1 1  10
  0 1 1 1 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 0  20
  0 1 0 0 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 1  20
  0 0 0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 0  50
  0 0 0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 1  50
  . 0 0 0 2 1 1 1 1 0  .16950354609929086 0 0 1 0 1 0  60
  0 1 0 0 1 1 0 1 1 0  .16950354609929086 0 0 1 0 1 1  60
  0 0 0 0 0 0 0 0 0 1  -.8304964539007091 0 1 0 0 1 0  70
  0 0 0 0 2 1 1 1 0 1  -.8304964539007091 0 1 0 0 1 1  70
  0 0 0 0 0 0 0 0 1 0  -.8304964539007091 1 0 1 0 1 0  90
  0 0 0 0 0 0 0 0 1 0  -.8304964539007091 1 0 1 0 1 1  90
  0 0 0 0 0 0 0 0 0 1 -1.8304964539007091 0 0 1 0 1 0 100
  0 0 0 0 0 0 0 0 0 1 -1.8304964539007091 0 0 1 0 1 1 100
  0 0 0 0 1 1 0 1 0 0  -2.830496453900709 0 0 1 0 1 0 140
  0 0 0 0 0 0 0 0 0 0  -2.830496453900709 0 0 1 0 1 1 140
  0 0 0 0 0 0 0 0 1 0  -2.830496453900709 0 0 0 0 1 0 150
  0 0 0 0 1 1 0 1 1 0  -2.830496453900709 0 0 0 0 1 1 150
  . 0 0 0 2 1 1 1 0 0  -.8304964539007091 0 0 1 1 1 0 170
  1 1 1 . 1 1 0 1 0 0  -.8304964539007091 0 0 1 1 1 1 170
  0 0 0 0 1 1 0 1 1 1  -.8304964539007091 0 0 1 0 1 0 200
  0 0 0 0 1 1 0 1 1 1  -.8304964539007091 0 0 1 0 1 1 200
  . 0 0 0 0 0 0 0 0 0  -.8304964539007091 0 0 0 0 1 0 210
  0 0 0 0 0 0 0 0 0 0  -.8304964539007091 0 0 0 0 1 1 210
  0 0 0 0 1 1 0 1 1 1 -1.8304964539007091 0 0 1 0 1 0 230
  0 0 0 0 1 1 0 1 1 1 -1.8304964539007091 0 0 1 0 1 1 230
  0 0 0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 0 240
  0 0 0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 1 240
  0 0 0 0 1 1 0 1 1 1  .16950354609929086 0 1 0 0 1 0 280
  0 0 0 0 0 0 0 0 1 1  .16950354609929086 0 1 0 0 1 1 280
  0 0 0 0 0 0 0 0 0 0 -1.8304964539007091 0 1 1 0 1 0 290
  0 0 0 0 1 1 0 0 0 0 -1.8304964539007091 0 1 1 0 1 1 290
  0 0 1 0 0 0 0 0 0 0   6.169503546099291 0 0 1 1 1 0 300
  0 0 1 0 0 0 0 0 0 0   6.169503546099291 0 0 1 1 1 1 300
  0 0 0 0 0 0 0 0 1 0   3.169503546099291 0 0 1 1 1 0 310
  0 0 0 0 1 1 0 0 1 0   3.169503546099291 0 0 1 1 1 1 310
  0 0 0 0 0 0 0 0 0 0  -.8304964539007091 0 1 0 0 1 0 340
  1 1 0 . 0 0 0 0 0 0  -.8304964539007091 0 1 0 0 1 1 340
  . 0 0 0 1 1 0 1 1 0   3.169503546099291 0 0 1 0 1 0 360
  0 0 0 0 1 1 0 1 1 0   3.169503546099291 0 0 1 0 1 1 360
  0 0 0 0 1 1 0 1 1 0  .16950354609929086 0 0 1 0 1 0 390
  0 0 0 0 1 1 0 0 1 0  .16950354609929086 0 0 1 0 1 1 390
  0 0 0 0 1 1 0 0 1 0 -1.8304964539007091 0 0 1 0 1 0 410
  0 0 0 0 1 1 0 1 1 0 -1.8304964539007091 0 0 1 0 1 1 410
  0 0 0 0 1 1 0 0 1 1  -3.830496453900709 0 0 0 0 1 0 440
  0 0 0 0 1 1 0 1 1 1  -3.830496453900709 0 0 0 0 1 1 440
  1 0 1 1 1 1 0 1 0 0  .16950354609929086 0 1 1 0 1 0 450
  . 1 0 . 1 1 0 1 0 0  .16950354609929086 0 1 1 0 1 1 450
  0 0 0 0 1 1 0 1 1 1  -.8304964539007091 0 0 1 0 1 0 470
  0 0 0 0 1 1 0 0 1 1  -.8304964539007091 0 0 1 0 1 1 470
  0 0 0 0 0 0 0 0 0 1  -3.830496453900709 1 1 1 0 1 0 480
  0 0 0 0 1 1 0 1 0 1  -3.830496453900709 1 1 1 0 1 1 480
  0 0 0 0 1 1 0 1 1 0   3.169503546099291 0 0 0 0 1 0 490
  0 1 0 0 1 1 0 1 1 0   3.169503546099291 0 0 0 0 1 1 490
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 0 0 1 1 0 510
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 0 0 1 1 1 510
  0 0 1 0 1 1 0 1 1 0  1.1695035460992909 0 0 0 1 1 0 530
  0 0 1 1 1 1 0 1 1 0  1.1695035460992909 0 0 0 1 1 1 530
  0 0 1 0 0 0 0 0 1 0  1.1695035460992909 0 0 1 0 1 0 540
  0 1 0 1 1 1 0 1 1 0  1.1695035460992909 0 0 1 0 1 1 540
  0 0 1 0 0 0 0 0 1 1   3.169503546099291 0 1 1 1 1 0 580
  0 1 1 1 1 1 0 1 1 1   3.169503546099291 0 1 1 1 1 1 580
  0 0 1 0 1 1 0 1 1 1   2.169503546099291 0 0 1 1 1 0 600
  0 0 0 0 1 1 0 1 1 1   2.169503546099291 0 0 1 1 1 1 600
  0 0 0 0 0 0 0 0 1 0  -.8304964539007091 0 0 1 0 1 0 620
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 0 1 0 1 1 620
  0 1 1 0 0 0 0 0 1 0  .16950354609929086 0 0 0 1 1 0 650
  0 1 0 0 1 1 0 1 1 0  .16950354609929086 0 0 0 1 1 1 650
  1 1 1 0 0 0 0 0 1 0  -.8304964539007091 0 0 0 0 1 0 680
  0 0 0 1 1 1 0 0 1 0  -.8304964539007091 0 0 0 0 1 1 680
  0 0 0 0 0 0 0 0 0 0  -3.830496453900709 0 1 1 0 1 0 690
  . 0 0 . 0 0 0 0 0 0  -3.830496453900709 0 1 1 0 1 1 690
  0 0 0 0 0 0 0 0 1 0  -2.830496453900709 0 1 0 0 1 0 700
  0 0 0 0 0 0 0 0 1 0  -2.830496453900709 0 1 0 0 1 1 700
  0 0 0 0 0 0 0 0 1 0  -3.830496453900709 0 0 0 0 1 0 710
  0 0 0 0 0 0 0 0 1 0  -3.830496453900709 0 0 0 0 1 1 710
  0 0 0 0 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 0 750
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 1 1 0 1 1 750
  0 0 0 0 2 1 1 1 1 1   2.169503546099291 0 0 1 1 1 0 780
  0 1 0 0 1 1 0 1 1 1   2.169503546099291 0 0 1 1 1 1 780
  . 0 0 0 1 1 0 1 0 0  1.1695035460992909 0 0 0 1 1 0 790
  0 1 0 0 1 1 0 1 0 0  1.1695035460992909 0 0 0 1 1 1 790
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 0 1 0 1 0 800
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 0 1 0 1 1 800
  . 0 0 0 0 0 0 0 1 0  1.1695035460992909 0 0 0 1 1 0 810
  0 0 0 0 1 1 0 1 1 0  1.1695035460992909 0 0 0 1 1 1 810
  . 0 0 0 1 1 0 1 1 1  .16950354609929086 0 1 0 0 1 0 830
  0 1 0 0 2 1 1 1 1 1  .16950354609929086 0 1 0 0 1 1 830
  0 0 0 0 1 1 0 0 1 0 -1.8304964539007091 0 1 0 0 1 0 850
  0 1 0 . 2 1 1 1 1 0 -1.8304964539007091 0 1 0 0 1 1 850
  0 0 0 0 0 0 0 0 1 1  -2.830496453900709 0 1 0 0 1 0 870
  0 0 0 0 1 1 0 0 1 1  -2.830496453900709 0 1 0 0 1 1 870
  0 0 0 0 2 1 1 1 1 0  1.1695035460992909 0 0 0 0 1 0 930
  0 0 0 0 1 1 0 0 1 0  1.1695035460992909 0 0 0 0 1 1 930
  0 0 0 0 0 0 0 0 1 0  -3.830496453900709 0 0 1 0 1 0 940
  0 0 0 0 1 1 0 0 1 0  -3.830496453900709 0 0 1 0 1 1 940
  . 0 0 0 0 0 0 0 1 0  .16950354609929086 1 0 0 1 1 0 950
  0 1 0 0 0 0 0 0 1 0  .16950354609929086 1 0 0 1 1 1 950
  0 0 0 0 0 0 0 0 1 1  -3.830496453900709 0 0 1 0 1 0 970
  0 0 0 0 0 0 0 0 1 1  -3.830496453900709 0 0 1 0 1 1 970
  end
  label values SexLastUnsafeConsist yes_no
  label values SexPartnerFreq2More yes_no
  label def yes_no 0 "0. No", modify
  label def yes_no 1 "1. Yes", modify
  label values Use UseT3
  label def UseT3 0 "No access", modify
  label def UseT3 1 "Social media", modify
  label def UseT3 2 "Health content", modify
  label values Sex Sex
  label def Sex 0 "Male", modify
  label def Sex 1 "Female", modify
  label values Rural UrbRur
  label def UrbRur 0 "0. Urban", modify
  label def UrbRur 1 "1. Rural", modify
  label values ARTever binary
  label values HomeTypeInformal binary
  label values Relationship binary
  label values SchEnrol binary
  label def binary 0 "0. No", modify
  label def binary 1 "1. Yes", modify
  label values NecessitiesAll_r binary_r
  label def binary_r 0 "0. Yes", modify
  label def binary_r 1 "1. No", modify
            The codes for the graph have not changed:

            Code:
            local y     SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex local x1    XC XM
  local xsex  Rural AgeC HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j
   
  foreach ys of local y {
      forvalues f = 0/1 {
          mi estimate, saving(`ys'`f', replace) cmdok : melogit `ys' `x1' `xsex' if Sex == `f' || ID:, or cov(un) 
          estimate store `ys'`f'
      }
  }
   
  coefplot    (SexDrunkDrugs0, label("Boy") msymbol(S) mcolor(black) mfcolor(white))      ///
  (SexDrunkDrugs1, label("Girl") msymbol(O) mcolor(black) mfcolor(black)), ///
  bylabel("Sex after substance use")  ///
  mlabel("{it:p} = " + string(@pval,"%9.3f")) mlabpos(1)                      ///
  || (SexLastUnsafeConsist0) (SexLastUnsafeConsist1),  ///
  bylabel("Infrequent condom use")                  ///
  || (SexPartnerFreq2More0)(SexPartnerFreq2More1),  ///
  bylabel("Multiple sexual partnership")         ///
  || (InequitSex0) (InequitSex1),  ///
  bylabel("Inequitable sexual partnership"),         ///
  || eform keep(XC XM) ciopts(lcolor(gs8) lwidth(thick)) ///
  xtitle("Adjusted odds ratios (ORs)") xlab(0(1)4,format(%3.0fc))  ///
  xline(1, lcolor(gs10)) ///
  ylab(1 `" "Health" "content" "use" "(Ref. no access)" "' 2 `" "Social" "media" "use" "(Ref. no access)" "') ///
  byopts(xrescale) subtitle(, fcolor(none) lstyle(none))  ///
  legend(order(2 "ORs for boys" 4 "ORs for girls" 3 "95% CI") row(1) ring(0) pos(10) region(lstyle(none)))

            Comment

            • #7
              Sorry, your data still is not sufficient to run any imputations. Please make sure you are able to reproduce the problem before posting the data and codes.

              Code:
              * Example generated by -dataex-. For more info, type help dataex
  clear
  input byte(SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More) float(InequitSex Use Access XC XM) byte(Sex Rural) double AgeC byte(HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j) long ID
  0 0 0 0 0 0 0 0 0 1  -.8304964539007091 1 1 1 0 1 0  10
  0 1 0 . 2 1 1 1 0 1  -.8304964539007091 1 1 1 0 1 1  10
  0 1 1 1 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 0  20
  0 1 0 0 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 1  20
  0 0 0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 0  50
  0 0 0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 1  50
  . 0 0 0 2 1 1 1 1 0  .16950354609929086 0 0 1 0 1 0  60
  0 1 0 0 1 1 0 1 1 0  .16950354609929086 0 0 1 0 1 1  60
  0 0 0 0 0 0 0 0 0 1  -.8304964539007091 0 1 0 0 1 0  70
  0 0 0 0 2 1 1 1 0 1  -.8304964539007091 0 1 0 0 1 1  70
  0 0 0 0 0 0 0 0 1 0  -.8304964539007091 1 0 1 0 1 0  90
  0 0 0 0 0 0 0 0 1 0  -.8304964539007091 1 0 1 0 1 1  90
  0 0 0 0 0 0 0 0 0 1 -1.8304964539007091 0 0 1 0 1 0 100
  0 0 0 0 0 0 0 0 0 1 -1.8304964539007091 0 0 1 0 1 1 100
  0 0 0 0 1 1 0 1 0 0  -2.830496453900709 0 0 1 0 1 0 140
  0 0 0 0 0 0 0 0 0 0  -2.830496453900709 0 0 1 0 1 1 140
  0 0 0 0 0 0 0 0 1 0  -2.830496453900709 0 0 0 0 1 0 150
  0 0 0 0 1 1 0 1 1 0  -2.830496453900709 0 0 0 0 1 1 150
  . 0 0 0 2 1 1 1 0 0  -.8304964539007091 0 0 1 1 1 0 170
  1 1 1 . 1 1 0 1 0 0  -.8304964539007091 0 0 1 1 1 1 170
  0 0 0 0 1 1 0 1 1 1  -.8304964539007091 0 0 1 0 1 0 200
  0 0 0 0 1 1 0 1 1 1  -.8304964539007091 0 0 1 0 1 1 200
  . 0 0 0 0 0 0 0 0 0  -.8304964539007091 0 0 0 0 1 0 210
  0 0 0 0 0 0 0 0 0 0  -.8304964539007091 0 0 0 0 1 1 210
  0 0 0 0 1 1 0 1 1 1 -1.8304964539007091 0 0 1 0 1 0 230
  0 0 0 0 1 1 0 1 1 1 -1.8304964539007091 0 0 1 0 1 1 230
  0 0 0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 0 240
  0 0 0 0 0 0 0 0 0 1  -3.830496453900709 0 1 1 0 1 1 240
  0 0 0 0 1 1 0 1 1 1  .16950354609929086 0 1 0 0 1 0 280
  0 0 0 0 0 0 0 0 1 1  .16950354609929086 0 1 0 0 1 1 280
  0 0 0 0 0 0 0 0 0 0 -1.8304964539007091 0 1 1 0 1 0 290
  0 0 0 0 1 1 0 0 0 0 -1.8304964539007091 0 1 1 0 1 1 290
  0 0 1 0 0 0 0 0 0 0   6.169503546099291 0 0 1 1 1 0 300
  0 0 1 0 0 0 0 0 0 0   6.169503546099291 0 0 1 1 1 1 300
  0 0 0 0 0 0 0 0 1 0   3.169503546099291 0 0 1 1 1 0 310
  0 0 0 0 1 1 0 0 1 0   3.169503546099291 0 0 1 1 1 1 310
  0 0 0 0 0 0 0 0 0 0  -.8304964539007091 0 1 0 0 1 0 340
  1 1 0 . 0 0 0 0 0 0  -.8304964539007091 0 1 0 0 1 1 340
  . 0 0 0 1 1 0 1 1 0   3.169503546099291 0 0 1 0 1 0 360
  0 0 0 0 1 1 0 1 1 0   3.169503546099291 0 0 1 0 1 1 360
  0 0 0 0 1 1 0 1 1 0  .16950354609929086 0 0 1 0 1 0 390
  0 0 0 0 1 1 0 0 1 0  .16950354609929086 0 0 1 0 1 1 390
  0 0 0 0 1 1 0 0 1 0 -1.8304964539007091 0 0 1 0 1 0 410
  0 0 0 0 1 1 0 1 1 0 -1.8304964539007091 0 0 1 0 1 1 410
  0 0 0 0 1 1 0 0 1 1  -3.830496453900709 0 0 0 0 1 0 440
  0 0 0 0 1 1 0 1 1 1  -3.830496453900709 0 0 0 0 1 1 440
  1 0 1 1 1 1 0 1 0 0  .16950354609929086 0 1 1 0 1 0 450
  . 1 0 . 1 1 0 1 0 0  .16950354609929086 0 1 1 0 1 1 450
  0 0 0 0 1 1 0 1 1 1  -.8304964539007091 0 0 1 0 1 0 470
  0 0 0 0 1 1 0 0 1 1  -.8304964539007091 0 0 1 0 1 1 470
  0 0 0 0 0 0 0 0 0 1  -3.830496453900709 1 1 1 0 1 0 480
  0 0 0 0 1 1 0 1 0 1  -3.830496453900709 1 1 1 0 1 1 480
  0 0 0 0 1 1 0 1 1 0   3.169503546099291 0 0 0 0 1 0 490
  0 1 0 0 1 1 0 1 1 0   3.169503546099291 0 0 0 0 1 1 490
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 0 0 1 1 0 510
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 0 0 1 1 1 510
  0 0 1 0 1 1 0 1 1 0  1.1695035460992909 0 0 0 1 1 0 530
  0 0 1 1 1 1 0 1 1 0  1.1695035460992909 0 0 0 1 1 1 530
  0 0 1 0 0 0 0 0 1 0  1.1695035460992909 0 0 1 0 1 0 540
  0 1 0 1 1 1 0 1 1 0  1.1695035460992909 0 0 1 0 1 1 540
  0 0 1 0 0 0 0 0 1 1   3.169503546099291 0 1 1 1 1 0 580
  0 1 1 1 1 1 0 1 1 1   3.169503546099291 0 1 1 1 1 1 580
  0 0 1 0 1 1 0 1 1 1   2.169503546099291 0 0 1 1 1 0 600
  0 0 0 0 1 1 0 1 1 1   2.169503546099291 0 0 1 1 1 1 600
  0 0 0 0 0 0 0 0 1 0  -.8304964539007091 0 0 1 0 1 0 620
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 0 1 0 1 1 620
  0 1 1 0 0 0 0 0 1 0  .16950354609929086 0 0 0 1 1 0 650
  0 1 0 0 1 1 0 1 1 0  .16950354609929086 0 0 0 1 1 1 650
  1 1 1 0 0 0 0 0 1 0  -.8304964539007091 0 0 0 0 1 0 680
  0 0 0 1 1 1 0 0 1 0  -.8304964539007091 0 0 0 0 1 1 680
  0 0 0 0 0 0 0 0 0 0  -3.830496453900709 0 1 1 0 1 0 690
  . 0 0 . 0 0 0 0 0 0  -3.830496453900709 0 1 1 0 1 1 690
  0 0 0 0 0 0 0 0 1 0  -2.830496453900709 0 1 0 0 1 0 700
  0 0 0 0 0 0 0 0 1 0  -2.830496453900709 0 1 0 0 1 1 700
  0 0 0 0 0 0 0 0 1 0  -3.830496453900709 0 0 0 0 1 0 710
  0 0 0 0 0 0 0 0 1 0  -3.830496453900709 0 0 0 0 1 1 710
  0 0 0 0 0 0 0 0 1 0  -.8304964539007091 0 1 1 0 1 0 750
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 1 1 0 1 1 750
  0 0 0 0 2 1 1 1 1 1   2.169503546099291 0 0 1 1 1 0 780
  0 1 0 0 1 1 0 1 1 1   2.169503546099291 0 0 1 1 1 1 780
  . 0 0 0 1 1 0 1 0 0  1.1695035460992909 0 0 0 1 1 0 790
  0 1 0 0 1 1 0 1 0 0  1.1695035460992909 0 0 0 1 1 1 790
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 0 1 0 1 0 800
  0 0 0 0 1 1 0 1 1 0  -.8304964539007091 0 0 1 0 1 1 800
  . 0 0 0 0 0 0 0 1 0  1.1695035460992909 0 0 0 1 1 0 810
  0 0 0 0 1 1 0 1 1 0  1.1695035460992909 0 0 0 1 1 1 810
  . 0 0 0 1 1 0 1 1 1  .16950354609929086 0 1 0 0 1 0 830
  0 1 0 0 2 1 1 1 1 1  .16950354609929086 0 1 0 0 1 1 830
  0 0 0 0 1 1 0 0 1 0 -1.8304964539007091 0 1 0 0 1 0 850
  0 1 0 . 2 1 1 1 1 0 -1.8304964539007091 0 1 0 0 1 1 850
  0 0 0 0 0 0 0 0 1 1  -2.830496453900709 0 1 0 0 1 0 870
  0 0 0 0 1 1 0 0 1 1  -2.830496453900709 0 1 0 0 1 1 870
  0 0 0 0 2 1 1 1 1 0  1.1695035460992909 0 0 0 0 1 0 930
  0 0 0 0 1 1 0 0 1 0  1.1695035460992909 0 0 0 0 1 1 930
  0 0 0 0 0 0 0 0 1 0  -3.830496453900709 0 0 1 0 1 0 940
  0 0 0 0 1 1 0 0 1 0  -3.830496453900709 0 0 1 0 1 1 940
  . 0 0 0 0 0 0 0 1 0  .16950354609929086 1 0 0 1 1 0 950
  0 1 0 0 0 0 0 0 1 0  .16950354609929086 1 0 0 1 1 1 950
  0 0 0 0 0 0 0 0 1 1  -3.830496453900709 0 0 1 0 1 0 970
  0 0 0 0 0 0 0 0 1 1  -3.830496453900709 0 0 1 0 1 1 970
  end
  label values SexLastUnsafeConsist yes_no
  label values SexPartnerFreq2More yes_no
  label def yes_no 0 "0. No", modify
  label def yes_no 1 "1. Yes", modify
  label values Use UseT3
  label def UseT3 0 "No access", modify
  label def UseT3 1 "Social media", modify
  label def UseT3 2 "Health content", modify
  label values Sex Sex
  label def Sex 0 "Male", modify
  label def Sex 1 "Female", modify
  label values Rural UrbRur
  label def UrbRur 0 "0. Urban", modify
  label def UrbRur 1 "1. Rural", modify
  label values ARTever binary
  label values HomeTypeInformal binary
  label values Relationship binary
  label values SchEnrol binary
  label def binary 0 "0. No", modify
  label def binary 1 "1. Yes", modify
  label values NecessitiesAll_r binary_r
  label def binary_r 0 "0. Yes", modify
  label def binary_r 1 "1. No", modify

    
   
  mi set mlong // mi set data for imputation
   
  mi reshape wide SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)
  
  
  * Identifying which variables in the imputation model have missing information.
  mi register imputed NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0 SexDrunkDrugs1 InequitSex0 InequitSex1
   
  *** Mutiple imputation
  *MI using chained equations/MICE (also known as the fully conditional specification or sequential generalized regression) for binary outcomes.  In simulation studies (Lee & Carlin, 2010; Van Buuren, 2007), MICE has been show to produce estimates that are comparable to MVN method.
   
  * Clear any xtset that is ongoing
  mi xtset, clear
  mi /*mimpt*/ impute chained  (logit) NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0  ///
  SexDrunkDrugs1 InequitSex0 InequitSex1 = Access0 Access1  ///
  XC0 XM0 XC1 XM1 Sex ARTever AgeC Rural Relationship  ///
  SchEnrol, add(10) rseed(20210101) /*skipnonconvergence(5)*/ savetrace(tracedata, replace)
   
  * Bring data back to long format
  mi reshape long SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)
   
  * We can check the data and see if there is no problems in our imputation
 summ SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex HomeTypeInformal NecessitiesAll_r
  
local y     SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex local x1    XC XM
  local xsex  Rural AgeC HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j
   
  foreach ys of local y {
      forvalues f = 0/1 {
          mi estimate, saving(`ys'`f', replace) cmdok : melogit `ys' `x1' `xsex' if Sex == `f' || ID:, or cov(un) 
          estimate store `ys'`f'
      }
  }
   
  coefplot    (SexDrunkDrugs0, label("Boy") msymbol(S) mcolor(black) mfcolor(white))      ///
  (SexDrunkDrugs1, label("Girl") msymbol(O) mcolor(black) mfcolor(black)), ///
  bylabel("Sex after substance use")  ///
  mlabel("{it:p} = " + string(@pval,"%9.3f")) mlabpos(1)                      ///
  || (SexLastUnsafeConsist0) (SexLastUnsafeConsist1),  ///
  bylabel("Infrequent condom use")                  ///
  || (SexPartnerFreq2More0)(SexPartnerFreq2More1),  ///
  bylabel("Multiple sexual partnership")         ///
  || (InequitSex0) (InequitSex1),  ///
  bylabel("Inequitable sexual partnership"),         ///
  || eform keep(XC XM) ciopts(lcolor(gs8) lwidth(thick)) ///
  xtitle("Adjusted odds ratios (ORs)") xlab(0(1)4,format(%3.0fc))  ///
  xline(1, lcolor(gs10)) ///
  ylab(1 `" "Health" "content" "use" "(Ref. no access)" "' 2 `" "Social" "media" "use" "(Ref. no access)" "') ///
  byopts(xrescale) subtitle(, fcolor(none) lstyle(none))  ///
  legend(order(2 "ORs for boys" 4 "ORs for girls" 3 "95% CI") row(1) ring(0) pos(10) region(lstyle(none)))
              Res.:

              Code:
              Conditional models:
    SexDrunkDrugs1: logit SexDrunkDrugs1 i.NecessitiesAll_r i.HomeTypeInformal i.InequitSex0 i.InequitSex1 i.SexDrunkDrugs0 Access0
                     Access1 XC0 XM0 XC1 XM1 Sex ARTever AgeC Rural Relationship SchEnrol
       InequitSex1: logit InequitSex1 i.NecessitiesAll_r i.HomeTypeInformal i.InequitSex0 i.SexDrunkDrugs1 i.SexDrunkDrugs0 Access0
                     Access1 XC0 XM0 XC1 XM1 Sex ARTever AgeC Rural Relationship SchEnrol
    SexDrunkDrugs0: logit SexDrunkDrugs0 i.NecessitiesAll_r i.HomeTypeInformal i.InequitSex0 i.SexDrunkDrugs1 i.InequitSex1 Access0
                     Access1 XC0 XM0 XC1 XM1 Sex ARTever AgeC Rural Relationship SchEnrol

Performing chained iterations ...
error occurred during imputation of SexDrunkDrugs0 SexDrunkDrugs1 InequitSex1 on m = 1
r(2000);

end of do-file
              • 1 like

              Comment

              • #8
                Thanks again Andrew. I realized that I had to increase the data rows for the imputation codes to work. However, Statalist has a limit in the number of characters to post (30000 I think). I will reach out if I find a way.

                Comment

                • #9
                  It's fine to upload the dataset in .dta format in this case.
                  • 1 like

                  Comment

                  • #10
                    I've used coefplot with MI data before, and this included labeling the dots with stars based on the p-value. I checked using my previous work. Using the mlabel option should work, e.g. using a code fragment like this one did correctly plot the p-value by the dot:

                    Code:
                    coefplot (SexDrunkDrugs0, label("Boy") msymbol(S) mcolor(black) mfcolor(white))... , mlabel(string(@pval,"%9.3f")) mlabpos(1)
                    Bolade, after you post the dta file, you could go back to your own code and simplify it. You've got a bunch of options added right now. Try plotting one model with just the mlabel option and see if it produces the p-value. Like Andrew said, I had used the post option with the original mi estimate: command; without it, I'm surprised that coefplot has anything to plot at all.

                    I've tried parsing the code and I can't see anything that would immediately explain why you're getting p-values plotted in MI, but not in regular estimation. The only possible thing I can see is that you placed the mlabel option with the first model you asked for, then you used the || operator and specified the other models. I have typically put mlabel as a global option after specifying all of the models to coefplot. I'm not sure why that would cause a failure with MI but not without it. Anyway, this is why I suggested you try plotting just one model to see if it changes anything.

                    This is a well-posed question. Thanks for including your complete code and a data extract. I hope we can resolve this. This has got to be resolvable, because I've done pretty much what you've done.
                    Last edited by Weiwen Ng; 24 Aug 2022, 15:24.
                    Be aware that it can be very hard to answer a question without sample data. You can use the dataex command for this. Type help dataex at the command line.

                    When presenting code or results, please use the code delimiters format them. Use the # button on the formatting toolbar, between the " (double quote) and <> buttons.
                    • 1 like

                    Comment

                    • #11
                      Originally posted by Andrew Musau View Post
                      It's fine to upload the dataset in .dta format in this case.
                      Thank you so much, Andrew! I am attaching the dataset.
                      Attached Files

                      Comment

                      • #12
                        Originally posted by Weiwen Ng View Post
                        I've used coefplot with MI data before, and this included labeling the dots with stars based on the p-value. I checked using my previous work. Using the mlabel option should work, e.g. using a code fragment like this one did correctly plot the p-value by the dot:

                        Code:
                        coefplot (SexDrunkDrugs0, label("Boy") msymbol(S) mcolor(black) mfcolor(white))... , mlabel(string(@pval,"%9.3f")) mlabpos(1)
                        Bolade, after you post the dta file, you could go back to your own code and simplify it. You've got a bunch of options added right now. Try plotting one model with just the mlabel option and see if it produces the p-value. Like Andrew said, I had used the post option with the original mi estimate: command; without it, I'm surprised that coefplot has anything to plot at all.

                        I've tried parsing the code and I can't see anything that would immediately explain why you're getting p-values plotted in MI, but not in regular estimation. The only possible thing I can see is that you placed the mlabel option with the first model you asked for, then you used the || operator and specified the other models. I have typically put mlabel as a global option after specifying all of the models to coefplot. I'm not sure why that would cause a failure with MI but not without it. Anyway, this is why I suggested you try plotting just one model to see if it changes anything.

                        This is a well-posed question. Thanks for including your complete code and a data extract. I hope we can resolve this. This has got to be resolvable, because I've done pretty much what you've done.
                        Thank you so much, Weiwen for your reply and your advice. I have tried your suggestion. Yet the same graph. The coefplot with p-value label showing indeed work perfectly with non-imputed data.
                        • 1 like

                        Comment

                        • #13
                          Thanks for the data example. I have cut out some of your twoway code for purpose of testing if the labels work. As I stated in #2, the issue that I see is that you do not add the -post- option to your mi estimate command. If I do so, I get the p-values printed.

                          Code:
                          use "20220825 sample data.dta", clear

  mi set mlong // mi set data for imputation
   
  mi reshape wide SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)
  
  
  * Identifying which variables in the imputation model have missing information.
  mi register imputed NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0 SexDrunkDrugs1 InequitSex0 InequitSex1
   
  *** Mutiple imputation
  *MI using chained equations/MICE (also known as the fully conditional specification or sequential generalized regression) for binary outcomes.  In simulation studies (Lee & Carlin, 2010; Van Buuren, 2007), MICE has been show to produce estimates that are comparable to MVN method.
   
  * Clear any xtset that is ongoing
  mi xtset, clear
  mi /*mimpt*/ impute chained  (logit) NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0  ///
  SexDrunkDrugs1 InequitSex0 InequitSex1 = Access0 Access1  ///
  XC0 XM0 XC1 XM1 Sex ARTever AgeC Rural Relationship  ///
  SchEnrol, add(10) rseed(20210101) /*skipnonconvergence(5)*/ savetrace(tracedata, replace)
   
  * Bring data back to long format
  mi reshape long SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)
   
  * We can check the data and see if there is no problems in our imputation
 summ SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex HomeTypeInformal NecessitiesAll_r
  
local y     SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex local x1    XC XM
  local xsex  Rural AgeC HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j
   
  foreach ys of local y {
      forvalues f = 0/1 {
          cap noi mi estimate, post saving(`ys'`f', replace) cmdok : melogit `ys' `x1' `xsex' if Sex == `f' || ID:, or cov(un) 
          cap noi estimate store `ys'`f'
      }
  }
   
  *set trace on
  set scheme s1color
  cap drop xpos
  gen xpos = 3.5
   coefplot    (SexDrunkDrugs0, label("Boy") msymbol(S) mcolor(black) mfcolor(white))      ///
  (SexDrunkDrugs1, label("Girl") msymbol(O) mcolor(black) mfcolor(black)), ///
  bylabel("Sex after substance use") mlabel(cond(@pval<0.001, "{it:p} <0.001", "{it:p} =" + string(@pval,"%9.3f"))) ///
    mlabc(none) addplot(scatter @at xpos, ms(i) mlab(@mlbl) mlabcolor(black) mlabpos(9) mlabgap(-10) mlabsize(small)) ///
    graphr(margin(r=10)) xsc(r(-10 5))
                          Click image for larger version Name: Graph.png Views: 1 Size: 29.4 KB ID: 1679202
                          • 1 like

                          Comment

                          • #14
                            Thanks a stack! I have adapted the code for another outcome
                            Code:
                            local y     SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex
local x0    Access
local x1     XC XM
local x     Rural AgeC Sex HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j
local xsex    Rural AgeC HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j
  foreach ys of local y {
      forvalues f = 0/1 {
          cap noi mi estimate, post saving(`ys'`f', replace) cmdok : melogit `ys' `x1' `xsex' if Sex == `f' || ID:, or cov(un) 
          cap noi estimate store `ys'`f'
      }
  }
    *set trace on
  set scheme s1color
  cap drop xpos
  gen xpos = 3.5
coefplot    (SexLastUnsafeConsist0, label("Boy") msymbol(S) mcolor(black) mfcolor(white))      ///
  (SexLastUnsafeConsist1, label("Girl") msymbol(O) mcolor(black) mfcolor(black)), ///
  bylabel("Infrequent condom use") mlabel(cond(@pval<0.001, "{it:p} <0.001", "{it:p} =" + string(@pval,"%9.3f"))) ///
    mlabc(none) addplot(scatter @at xpos, ms(i) mlab(@mlbl) mlabcolor(black) mlabpos(9) mlabgap(-10) mlabsize(small)) ///
    graphr(margin(r=10)) xsc(r(-10 5))
                            , and this is get the p-values missing. I got the following (attached) graph.

                            I looked at the regression tables and I can't see what might be the problem.
                            Attached Files
                            Last edited by Bolade Banougnin; 25 Aug 2022, 07:16.

                            Comment

                            • #15
                              Please post images in .PNG format as recommended in the FAQs. I am not able to open the .GPH image created by Stata 17 in Stata 16. That being the case, I ran the code in #14 and three p-values were missing.
                              Click image for larger version Name: Graph1.png Views: 1 Size: 43.2 KB ID: 1679353




                              On further investigation, I note that these values are not missing in the stored results (highlighted below).

                              Code:
                              . est replay SexLastUnsafeConsist0

--------------------------------------------------------------------------------------
Model SexLastUnsafeConsist0
--------------------------------------------------------------------------------------

Mixed-effects logistic regression               Number of obs     =      1,174
Group variable:              ID                 Number of groups  =        587

                                                Obs per group:
                                                              min =          2
                                                              avg =        2.0
                                                              max =          2

Integration method: mvaghermite                 Integration pts.  =          7

                                                Wald chi2(.)      =          .
Log likelihood =          .                     Prob > chi2       =          .
-------------------------------------------------------------------------------------
SexLastUnsafeCons~t |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
--------------------+----------------------------------------------------------------
                 XC |  -.6959997   .6446828    -1.08   0.280    -1.959555    .5675554
                 XM |   .3778831   .3858202     0.98   0.327    -.3783106    1.134077
              Rural |  -.6248963   .5701184    -1.10   0.273    -1.742308    .4925153
               AgeC |   .4053572   .1053465     3.85   0.000     .1988818    .6118326
   HomeTypeInformal |  -.5606224   .7523278    -0.75   0.456    -2.035158    .9139131
   NecessitiesAll_r |   .5613208    .474195     1.18   0.237    -.3680843    1.490726
            ARTever |  -.8648473       .479    -1.81   0.071     -1.80367    .0739755
       Relationship |   2.024283   .5228796     3.87   0.000      .999458    3.049108
           SchEnrol |  -2.499274   1.336178    -1.87   0.061    -5.118134    .1195862
                  j |   .8185097   .2963594     2.76   0.006     .2376559    1.399363
              _cons |  -2.661133   1.376298    -1.93   0.053    -5.358626    .0363608
--------------------+----------------------------------------------------------------
      var(_cons[ID])|   8.865928   3.155539                      4.413333    17.81073
-------------------------------------------------------------------------------------
LR test vs. logistic model: chi2(.) = .                   Prob > chi2 =      .

Note: LR test is conservative and provided only for reference.

. est replay SexLastUnsafeConsist1

--------------------------------------------------------------------------------------
Model SexLastUnsafeConsist1
--------------------------------------------------------------------------------------

Mixed-effects logistic regression               Number of obs     =      1,532
Group variable:              ID                 Number of groups  =        766

                                                Obs per group:
                                                              min =          2
                                                              avg =        2.0
                                                              max =          2

Integration method: mvaghermite                 Integration pts.  =          7

                                                Wald chi2(.)      =          .
Log likelihood =          .                     Prob > chi2       =          .
-------------------------------------------------------------------------------------
SexLastUnsafeCons~t |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
--------------------+----------------------------------------------------------------
                 XC |  -.6728635   .3554466    -1.89   0.058    -1.369526     .023799
                 XM |   .3563452    .215419     1.65   0.098    -.0658684    .7785587
              Rural |   .2718676    .235966     1.15   0.249    -.1906173    .7343525
               AgeC |   .5020056    .061067     8.22   0.000     .3823165    .6216948
   HomeTypeInformal |   .1759705   .2726909     0.65   0.519    -.3584938    .7104348
   NecessitiesAll_r |  -.0812407   .2505907    -0.32   0.746    -.5723895    .4099081
            ARTever |   -.670258   .2284569    -2.93   0.003    -1.118025   -.2224908
       Relationship |   1.025744   .2646242     3.88   0.000     .5070905    1.544398
           SchEnrol |   .1936175   .3525943     0.55   0.583    -.4974547    .8846896
                  j |    1.00031   .1776544     5.63   0.000      .652114    1.348506
              _cons |    -3.3551   .4931144    -6.80   0.000    -4.321587   -2.388614
--------------------+----------------------------------------------------------------
      var(_cons[ID])|   2.449458   .6544243                      1.450954    4.135104
-------------------------------------------------------------------------------------
LR test vs. logistic model: chi2(.) = .                   Prob > chi2 =      .

Note: LR test is conservative and provided only for reference.
                              Furthermore, if I asked for the coefficients to be printed, they are available. Since you have 1500+ observations, my workaround involves approximating the p-values using the normal distribution in these cases where they are missing. I can only conclude that this is a bug in coefplot. I would recommend that you send a dataset and some code that reproduces the issue to Ben Jann so that he may have a look at it. The approximation is highlighted in the code below.

                              Code:
                               use "C:\Users\amus\Downloads\20220825 sample data.dta"

  mi set mlong // mi set data for imputation
  
  mi reshape wide SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)
  
  
  * Identifying which variables in the imputation model have missing information.
  mi register imputed NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0 SexDrunkDrugs1 InequitSex0 InequitSex1
  
  *** Mutiple imputation
  *MI using chained equations/MICE (also known as the fully conditional specification or sequential generalized regression) for binary outcomes.  In simulation studies (Lee & Carlin, 2010; Van Buuren, 2007), MICE has been show to produce estimates that are comparable to MVN method.
  
  * Clear any xtset that is ongoing
  mi xtset, clear
  mi /*mimpt*/ impute chained  (logit) NecessitiesAll_r HomeTypeInformal SexDrunkDrugs0  ///
  SexDrunkDrugs1 InequitSex0 InequitSex1 = Access0 Access1  ///
  XC0 XM0 XC1 XM1 Sex ARTever AgeC Rural Relationship  ///
  SchEnrol, add(10) rseed(20210101) /*skipnonconvergence(5)*/ savetrace(tracedata, replace)
  
  * Bring data back to long format
  mi reshape long SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex Access XC XM Use, i(ID) j(j)
  
  * We can check the data and see if there is no problems in our imputation
 summ SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex HomeTypeInformal NecessitiesAll_r
local y     SexDrunkDrugs SexLastUnsafeConsist SexPartnerFreq2More InequitSex
local x0    Access
local x1     XC XM
local x     Rural AgeC Sex HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j
local xsex    Rural AgeC HomeTypeInformal NecessitiesAll_r ARTever Relationship SchEnrol j
  foreach ys of local y {
      forvalues f = 0/1 {
          cap noi mi estimate, post saving(`ys'`f', replace) cmdok : melogit `ys' `x1' `xsex' if Sex == `f' || ID:, or cov(un)
          cap noi estimate store `ys'`f'
      }
  }
    *set trace on
  set scheme s1color
  cap drop xpos
  gen xpos = 3.5
coefplot (SexLastUnsafeConsist0, label("Boy") msymbol(S) mcolor(black) mfcolor(white))      ///
  (SexLastUnsafeConsist1, label("Girl") msymbol(O) mcolor(black) mfcolor(black)), ///
  bylabel("Infrequent condom use") mlabel(cond(@pval<0.001, "{it:p} < 0.001", ///
  cond(@pval>0.001 &@pval<., "{it:p} = " + string(@pval,"%9.3f"), ///
  cond(missing(@pval) & 2*normal(@t)<0.001, "{it:p} < 0.001", "{it:p} = " + string(2*normal(@t),"%9.3f")) ))) ///
  mlabc(none) addplot(scatter @at xpos, ms(i) mlab(@mlbl) mlabcolor(black) ///
  mlabpos(9) mlabgap(-10) mlabsize(small)) graphr(margin(r=10)) xsc(r(-10 5))
                              Click image for larger version Name: Graph2.png Views: 1 Size: 33.5 KB ID: 1679354

                              Last edited by Andrew Musau; 26 Aug 2022, 02:50.
                              • 2 likes

                              Comment

                              Previous 1 2 Next
                              Working...
                              X