Hello there! I would like to ask the following error when using the proc nlmixed process for cross-layer interaction testing when doing a two-level logistic regression model: ERROR: Quadrature accuracy of 0.000100 could not be achieved with 31 points. The achieved accuracy was 1.000000. How should I do? My level 2 explanatory variables are multi-class variables. The sas code is as follows: data xb; d_0=h_0*entity; d_2=h_2*entity; d_3=h_3*entity; proc nlmixed; parms B0=-1.3365 B_g=0.4312 B_age=-0.09340 B_e=1 B_o0=0.4114,B_o2=-1.6674,B_o3=0.2509,B_o4=3.1463 B_h0=0.5347,B_h2=0.7353,B_h3=0.1027,V0=0.6416,v1=0,cov=1 B_d0=1 B_d2=1 B_d3=1; z=B0+B_g*gender+B_age*age1+B_e*entity+B_o0*o_0+B_o2*o_2+B_o3*o_3+B_o4*o_4+B_h0*h_0+B_h2*h_2+B_h3*h_3 +B_d0*d_0+B_d2*d_2+B_d3*d_3+u0j+u1j*entity; if(result=1) then p=1/(1+exp(-z)); else p=1-(1/(1+exp(-z))); LL=log(p); Model result~general(P);*定义结局变量及其分布; random u0j u1j~normal([0,0],[v0,cov,v1])subject=doctor; estimate 'Var(u)' V0; estimate 'Var(u1)' V1; run; Is my code wrong?
... View more