Open 11zhai opened 1 year ago
Hi 11zhai, thank you for your interest. This is because we presented a simplified example of drawing Figure 4. The min95 & max95 are obtained from "reghdfe" regression (just above the plot code), and you need to record them for each heterogeneity regression, i.e., for each subgroup analysis. They will be the elements of the plot.
Thanks
graph twoway (rspike -26 -13 order, horizontal lwidth(medthin) lcolor(black%50) lpattern(dash) lcolor(black%40)) /// (scatter order estimate, msize(medsmall) mcolor("0 183 255") msymbol(diamond)) /// ,scheme(s2color) /// xsize(5) ysize(6) /// ylabel(1 "Dust Emission (L)" 1.5 "Dust Emission (H)" /// 2 "SO{sub:2} Emission (L)" 2.5 "SO{sub:2} Emission (H)" /// 3 "Wastewater Emission (L)" 3.5 "Wastewater Emission (H)" /// 5 "Amount of Traffic (L)" 5.5 "Amount of Traffic (H)" /// 6 "# of Firms (L)" 6.5 "# of Firms (H)" /// 7 "Secondary Industry Output (L)" 7.5 "Secondary Industry Output (H)" /// 5 "Amount of Traffic (L)" 5.5 "Amount of Traffic (H)" /// 9 "Population (L)" 9.5 "Population (H)" /// 10 "per capita GDP (L)" 10.5 "per capita GDP (H)" /// 11 "GDP (L)" 11.5 "GDP (H)" /// 13 "Southern China" 13.5 "Northern China" /// 14 "Warm Region" 14.5 "Cold Region" /// 16 "Baseline" /// , labsize(small)) /// ytitle("") /// xtitle("Estimated Coefficient", size(small)) /// xlabel(, labsize(small)) /// xline(0, lpattern(dot)) /// yline(4.25 8.25 12.25 15.25, lpattern(dash) lcolor(orange%15) lwidth(thick)) /// legend(off) /// text(14.9 -0.4 "{it:Panel A}", place(e) size(2.5)) /// text(11.9 -0.4 "{it:Panel B}", place(e) size(2.5)) /// text(7.9 -0.4 "{it:Panel C}", place(e) size(2.5)) /// text(3.9 -0.4 "{it:Panel D}", place(e) size(2.5))
_Originally posted by @11zhai in https://github.com/yhyhpan/COVID19_LOCKDOWN/issues/5#issuecomment-1359384497_