Hi all, Behavior of restrictions with patterns like "b[1, 1] + -2*b[2, 1] = 0" is strange. In the following MWE, the restricted coefficients differ by exactly 1.0 when restrictions are defined in the "bugged" way: open greene13_1.gdt # set up system system name="Grunfeld" equation I_GM const C_GM F_GM equation I_WE const C_WE F_WE end system # restriction w/bug restrict "Grunfeld" b[1,1] + -1.0*b[2,1] = 0.0 end restrict estimate "Grunfeld" method=sur --iterate # restriction w/o bug restrict "Grunfeld" b[1,1] - 1.0*b[2,1] = 0.0 end restrict estimate "Grunfeld" method=sur -iterate This a bug or a feature? Thanks for looking into this. Best, Igor Igor Pozdeev +41786120928 www.igorpozdeev.me<http://www.igorpozdeev.me/>