Code gives O/P even if features are perfectly collinear in linear regression

Question

Columns 1 and 2 are perfectly correlated to each other by linear function c1+c2=10, but then also the code gives O/P.How is it possible, shouldn\'t the matrix b