以下是命令，平行趋势检验的系数与基准回归中did交互项的系数相反请问该如何解释呢？
tsset code year
gen did= treat* post
gen event=year - 2013 if treat==1
tab event, gen(eventz)
forvalues i = 1/17{
    replace eventz`i' = 0 if eventz`i' == .
}
reghdfe lnemission eventz*  PER density IS EDU ene1 t , absorb(treat post) vce(cluster code)
forvalue i=2/8{
    gen b_`i' = _b[eventz`i']
}
gen avg_coef=(b_2+b_3+b_4+b_5+b_6+b_7+b_8)/7
su avg_coef
return list
coefplot, baselevels keep(eventz*) coeflabels(eventz2 = "-7" eventz3 = "-6" eventz4 = "-5" eventz5 = "-4" eventz6 = "-3" eventz7 = "-2" eventz8 = "-1" eventz9 = "current" eventz10 = "1" eventz11 = "2" eventz12 = "3" eventz13 = "4"  eventz14 = "5" eventz15 = "6" eventz16 = "7" eventz17 = "8") vertical yline(0, lwidth(vthin) lpattern(dash) lcolor(teal)) ylabel(-0.5(0.2)1.5) xline(7, lwidth(vthin) lpattern(dash) lcolor(teal)) ytitle("处理效应系数估计值", size(small)) xtitle("时间(横坐标为完整T、去除系数均值后画图)", size(small)) transform(*=@-r(mean)) addplot(line @b @at) ciopts(lpattern(dash) recast(rcap) msize(medium)) msymbol(circle_hollow) scheme(s1mono)