MLR regression error "near singular matrix"

[alanlamtat]

Dear all, I am currently doing a MLR regression using eViews. There are two independent variables, namely X1 and X2 and are defined as follows:
X1 = 10 + Ux1
X2 = 40
Y = 8*X1 + 6*X2 + 10*Uy
When I regress Y on X1 and X2, i.e. Y c X1 X2, there is an error message from eViews which says near singular matrix. May somebody explain why this happen please? Does this relate to and violate any assumptions made in MLR? Thanks.

[startz]

Dear all, I am currently doing a MLR regression using eViews. There are two independent variables, namely X1 and X2 and are defined as follows:
X1 = 10 + Ux1
X2 = 40
Y = 8*X1 + 6*X2 + 10*Uy
When I regress Y on X1 and X2, i.e. Y c X1 X2, there is an error message from eViews which says near singular matrix. May somebody explain why this happen please? Does this relate to and violate any assumptions made in MLR? Thanks.

X2 is a constant.

[alanlamtat]

Yep, thanks for your reply. Would you mind giving me further explanations by referencing some sorts of formulae or assumptions which are linked to multiple linear regression? Greatly appreciate your efforts. Thanks.

[startz]

Take a look at Wikipedia, or most any econometrics text.