Overlay two coefplots

[Zariab Hossain]

Hi Stata users,

I am trying to overlay two coefplots but I am having some difficulty. Here is my command:

keep if l2h==1

summ     event_time
    g         t = event_time - r(min)
    summ     t if event_time == -1
    local     true_neg = r(mean)

reghdfe    breadth_search c.ra_delta##ib`true_neg'.t ///
                LWbas_male age LWbas_married LWbas_immigrant LWbas_child LWbas_pre LWbas_pre9 ///
                LWbas_s_high LWbas_l_high LWbas_s_univ LWbas_l_univ LW5_emp_mon_last5y ///
                LWhist_empbef1 LW5_av_dur LW5_last_dur LWhist_income_bef1 LW5_welfare5, ///
                cluster(referal_officeid) absorb(referal_year referal_CWid llm_kod future_officeid) 
    estimates store l2h

    coefplot (l2h, msymbol(Sh) msize(small) label(RA_Spillover_l2h)), ///
    keep(10.t#c.ra_delta 11.t#c.ra_delta 12.t#c.ra_delta ///
    13.t#c.ra_delta 14.t#c.ra_delta 15.t#c.ra_delta 16.t#c.ra_delta 17.t#c.ra_delta 18.t#c.ra_delta 19.t#c.ra_delta 20.t#c.ra_delta) ///
    rename(10.t#c.ra_delta= "t=-4" 11.t#c.ra_delta= "t=-3" 12.t#c.ra_delta= "t=-2" ///
    13.t#c.ra_delta ="t=-1" 14.t#c.ra_delta= "t=0" 15.t#c.ra_delta= "t=1" 16.t#c.ra_delta ="t=2" 17.t#c.ra_delta ="t=3" 18.t#c.ra_delta ="t=4" 19.t#c.ra_delta ="t=5" 20.t#c.ra_delta ="t=6") ///
    omitted baselevels xline(4, lcolor(gs5) lp(dash) lwidth(medium)) ///
    yline(0, lcolor(black) lp(dash) lwidth(medium))  ///
    legend(size(vsmall)) ylabel(-.8(.4).8, labsize(vsmall)) xlabel(, labsize(vsmall)) noci recast(connected) ciopts(recast(rcap)) lcolor(red) lp(dash) vertical

keep if l2h==0

summ     event_time
    g         t = event_time - r(min)
    summ     t if event_time == -1
    local     true_neg = r(mean)
    
    reghdfe        breadth_search c.ra_delta##ib`true_neg'.t ///
                LWbas_male age LWbas_married LWbas_immigrant LWbas_child LWbas_pre LWbas_pre9 ///
                LWbas_s_high LWbas_l_high LWbas_s_univ LWbas_l_univ LW5_emp_mon_last5y ///
                LWhist_empbef1 LW5_av_dur LW5_last_dur LWhist_income_bef1 LW5_welfare5, ///
                cluster(referal_officeid) absorb(referal_year referal_CWid llm_kod future_officeid) 
    estimates store h2l
    
    coefplot (h2l, msymbol(Sh) msize(small) label(RA_Spillover_h2l)), ///
    keep(9.t#c.ra_delta 10.t#c.ra_delta 11.t#c.ra_delta 12.t#c.ra_delta ///
    13.t#c.ra_delta 14.t#c.ra_delta 15.t#c.ra_delta 16.t#c.ra_delta 17.t#c.ra_delta 18.t#c.ra_delta 19.t#c.ra_delta) ///
    rename(9.t#c.ra_delta ="t=-4" 10.t#c.ra_delta= "t=-3" 11.t#c.ra_delta= "t=-2" 12.t#c.ra_delta= "t=-1" ///
    13.t#c.ra_delta ="t=0" 14.t#c.ra_delta= "t=1" 15.t#c.ra_delta= "t=2" 16.t#c.ra_delta ="t=3" 17.t#c.ra_delta ="t=4" 18.t#c.ra_delta ="t=5" 19.t#c.ra_delta ="t=6") ///
    omitted baselevels xline(4, lcolor(gs5) lp(dash) lwidth(medium)) ///
    yline(0, lcolor(black) lp(dash) lwidth(medium))  ///
    legend(size(vsmall)) ylabel(-.8(.4).8, labsize(vsmall)) xlabel(, labsize(vsmall)) noci recast(connected) ciopts(recast(rcap)) lcolor(red) lp(dash) vertical

(i) I also want to draw a common xline at 0 but this point is defined differently in these two models.
(ii) I do not want to connect the coefficients between time=-1 and time=0.

Any suggestion or help would be highly appreciated.

Best,
Zariab Hossain
Uppsala University

[Zariab Hossain]

Hi ,

I have come up with the following code that helped me to overlay the above coefplots. I do not need confidence interval but I want to place the coefficients one the top of each other (NOT side by side!). I want all the point estimates to be connected except between t=-1 and t=0. How can I do that?

Here is the code:

    coefplot (l2h, keep(10.t#c.ra_delta 11.t#c.ra_delta 12.t#c.ra_delta 13.t#c.ra_delta 14.t#c.ra_delta 15.t#c.ra_delta 16.t#c.ra_delta 17.t#c.ra_delta 18.t#c.ra_delta 19.t#c.ra_delta 20.t#c.ra_delta) rename(10.t#c.ra_delta= "t=-4" 11.t#c.ra_delta= "t=-3" 12.t#c.ra_delta= "t=-2" 13.t#c.ra_delta ="t=-1" 14.t#c.ra_delta= "t=0" 15.t#c.ra_delta= "t=1" 16.t#c.ra_delta ="t=2" 17.t#c.ra_delta ="t=3" 18.t#c.ra_delta ="t=4" 19.t#c.ra_delta ="t=5" 20.t#c.ra_delta ="t=6")) (h2l, keep(9.t#c.ra_delta 10.t#c.ra_delta 11.t#c.ra_delta 12.t#c.ra_delta 13.t#c.ra_delta 14.t#c.ra_delta 15.t#c.ra_delta 16.t#c.ra_delta 17.t#c.ra_delta 18.t#c.ra_delta 19.t#c.ra_delta) rename(9.t#c.ra_delta ="t=-4" 10.t#c.ra_delta= "t=-3" 11.t#c.ra_delta= "t=-2" 12.t#c.ra_delta= "t=-1" 13.t#c.ra_delta ="t=0" 14.t#c.ra_delta= "t=1" 15.t#c.ra_delta= "t=2" 16.t#c.ra_delta ="t=3" 17.t#c.ra_delta ="t=4" 18.t#c.ra_delta ="t=5" 19.t#c.ra_delta ="t=6")), yline(0, lcolor(black) lp(dash) lwidth(medium)) noci vertical xline(4) ylabel(-.8(.4).8, labsize(vsmall)) xlabel(, labsize(vsmall))

Here is the plot:



Best,
Zariab Hossain

[Zariab Hossain]

I solved the problem.

Best,
Zariab Hossain