1: \begin{abstract}

2: The density matrix renormalization group (DMRG) algorithm was originally designed

3: to efficiently compute the zero temperature or ground-state properties of one 

4: dimensional strongly correlated quantum systems. The development of the algorithm

5: at finite temperature has been a topic of much interest, because of the usefulness

6: of thermodynamics quantities in understanding the physics of condensed matter

7: systems, and because of the increased complexity associated with efficiently computing 

8: temperature-dependent properties. The ancilla method is a DMRG technique that enables

9: the computation of these thermodynamic quantities. In this paper, we review the

10: ancilla method, and improve its performance by working on reduced Hilbert spaces and

11: using canonical approaches. We furthermore

12: explore its applicability beyond spins systems to t-J and Hubbard models.

13: \end{abstract}

14: