<font color="#000000" size="2" face="Verdana"><div><font size="2" face="Verdana"><div><font size="2" face="Verdana"><div>Dear all,</div>
<div></div>
<div> </div>
<div>Hello, everyone. We are two senior students in USTC, China. </div>
<div></div>
<div align="left"> </div>
<div align="left">We encounted some problems in derivation of formula(34) in "Maximally localized generalized </div><div align="left">Wannier functions for composite energy bands", written by N.Marzari and D.Vanderbilt,</div>
<div align="left"> 1997. In this paper, it says that formula (34) was derived partly using formula(32), however, </div><div align="left">we could not get rid of the last term "ImlnM_{nn}^{k,b}" originally in formula (32). </div>
<div align="left"></div>
<div align="left"> </div>
<div align="left">The second question is that it is difficult for me to obtain formula (26). It seems using Taylor </div><div align="left">expansion to expand M_{nn}^{k,b} in orders of b. However, I find it not so straightforward. </div>
<div align="left"></div>
<div align="left"> </div>
<div align="left">The
third question comes from "APPENDIX B: FINITE-DIFFERENCE FORMULAS FOR k-SPACE GRIDS". We want to know where can I find the original theory of finitre-difference that would leads to the "B1 condition". Where did the concept of using </div><div align="left">weight factor "w_{b}" come from? What is the role it plays in </div>
<div align="left">finite-difference for k grids?</div>
<div align="left"> </div>
<div align="left"></div>
<div align="left">Thanks a lot.</div>
<div> </div>
<div></div>
<div>Shipeng Shu, Chen Liao</div></font></div></font></div></font>