Dear Stata users,

I am using the user-written command coefplot to plot coefficients from 27 models in subgraphs organized by outcome variables. Please see the code and output below. My question is how can I supress the variables (and corresponding spaces) that do not have markers from the subgraphs?

Code:
#delimit ;
coefplot (HH_ACTxT15, pstyle(p1)) (HH_ACTxT16, pstyle(p2)) (HH_ACTxT17, pstyle(p3)), bylabel(example 1) subtitle(, size(small))  keep (Netw_HH_ACT Netw_HH_ACT_promoter Netw_HH_ACT_lead1) ||
(DACTxT15, pstyle(p1)) (DACTxT16, pstyle(p2)) (DACTxT17, pstyle(p3)) , bylabel(example 2) subtitle(, size(small))  keep (Netw_HH_ACT Netw_HH_ACT_promoter Netw_HH_ACT_lead1)  ||
(NacthhmxT15, pstyle(p1)) (NacthhmxT16, pstyle(p2)) (NacthhmxT17, pstyle(p3)), bylabel(example 3) subtitle(, size(small))  keep (Netw_HH_ACT Netw_HH_ACT_promoter Netw_HH_ACT_lead1)  ||
(HH_ACTxT18, pstyle(p1)) (HH_ACTxT19, pstyle(p2)) (HH_ACTxT20, pstyle(p3)), bylabel(example 1) subtitle(, size(small))  keep (Netw_DACT Netw_DACT_promoter Netw_DACT_lead1) ||
(DACTxT18, pstyle(p1)) (DACTxT19, pstyle(p2)) (DACTxT20, pstyle(p3)), bylabel(example 2) subtitle(, size(small))  keep (Netw_DACT Netw_DACT_promoter Netw_DACT_lead1) ||
(NacthhmxT18, pstyle(p1)) (NacthhmxT19, pstyle(p2)) (NacthhmxT20, pstyle(p3)), bylabel(example 3) subtitle(, size(small))  keep (Netw_DACT Netw_DACT_promoter Netw_DACT_lead1) ||
(HH_ACTxT21, pstyle(p1)) (HH_ACTxT22, pstyle(p2)) (HH_ACTxT23, pstyle(p3)), bylabel(example 1) subtitle(, size(small))  keep (Netw_Nacthhm Netw_Nacthhm_promoter Netw_Nacthhm_lead1) ||
(DACTxT21, pstyle(p1)) (DACTxT22, pstyle(p2)) (DACTxT23, pstyle(p3)), bylabel(example 2) subtitle(, size(small))  keep (Netw_Nacthhm Netw_Nacthhm_promoter Netw_Nacthhm_lead1) ||  
(NacthhmxT21, pstyle(p1)) (NacthhmxT22, pstyle(p2)) (NacthhmxT23, pstyle(p3)), bylabel(example 3) subtitle(, size(small)) keep (Netw_Nacthhm Netw_Nacthhm_promoter Netw_Nacthhm_lead1)
xline(0, lcolor(black) lpattern(dash)) byopts(row(3) xrescale) nokey plotregion(margin(medium)) nooffsets
;

Many thanks,
Rute