1: \begin{abstract}

2: We develop coarse-graining tensor renormalization group algorithms to

3: compute physical properties of two-dimensional lattice models on finite

4: periodic lattices. Two different coarse-graining strategies, one based on

5: the tensor renormalization group and the other based on the higher-order

6: tensor renormalization group, are introduced. In order to optimize the

7: tensor-network model globally, a sweeping scheme is proposed to account for

8: the renormalization effect from the environment tensors under the framework

9: of second renormalization group. We demonstrate the algorithms by the

10: classical Ising model on the square lattice and the Kitaev model on the

11: honeycomb lattice, and show that the finite-size algorithms achieve

12: substantially more accurate results than the corresponding infinite-size

13: ones.

14: \end{abstract}

15: