Normalized image of a vector by an infinite product of nonnegative matrices
To prove that a measure, linearly representable by means of a finite set of nonnegative matrices $\mathcal M$, has the weak-Gibbs property, one check the uniform convergence (on $\mathcal M^\mathbb N$) of the sequence of vectors $\frac{A_1\cdots A_nc}{\Vert A_1\cdots A_nc\Vert}$ ($c$ positive column-vector). The main theorem gives a sufficient condition for this sequence to converge pointwise. This theorem generalizes the Birkhoff contraction method because it can be used even if the matrices have many zero entries. We also look at the convergence of the sequence of matrices $\frac{A_1\cdots A_n}{\Vert A_1\cdots A_n\Vert}$. The measures defined by Bernoulli convolution are in certain cases linearly representable; we give two example of weak-Gibbs Bernoullt convolutions, by using the Birkhoff contraction coefficient for the first and the theorem for the second. Furthermore we explicit the relationship between the notions of Bernoulli convolution, fundamental curves and lattice two-scale difference equations.