August 2007 - Comp-phys-alps-users

DMRG question
by Justin David Peel 11 Dec '07

11 Dec '07

Hi all, I just have a quick question. How can I get the dmrg program to output the max truncation error during the final sweep (or any truncation error for that matter) to the output file? I know that it is output to the screen, but I would like it in the output file. Thanks, Justin

2 1

one question about complex parameter
by GANG CHEN 28 Aug '07

28 Aug '07

Hi, all The site operators for spin are defined as S+, S- and Sz in ALPS, in order to express Sy, we need to introduce "i". Now I have a spin Hamiltonian, if I express it in terms of S+, S- and Sz, there will be "i" terms in the Hamiltonian. In such case, how do I specify each term or parameters? Thanks, GANG CHEN

2 1

two questions
by GANG CHEN 23 Aug '07

23 Aug '07

Hi all, In the lattice I am working on, there are N (N>2) equivalent sites in primitive unit cell. All the bonds are equivalent, but have different orientation. e.g. kagome lattice, the three bonds have different directions. Now I need to write down the Dzyloshiski-Moriya interaction for this lattice, the D vector depends on the direction of the bond, or more specifically, D vector depends on the relative coordinates of the two sites. My question is how to specify the bond operator for different bonds, or, how to specify the coordinate-dependence of the bond operators? The second question is about how to specify unit cell of the lattice. Here are two concrete examples from the library. I think if i can understand them, i won't have trouble with my lattice. 1. triangular, <UNITCELL name="triangular" dimension="2"> <VERTEX/> <EDGE><SOURCE vertex="1" offset="0 0"/><TARGET vertex="1" offset="0 1"/></EDGE> <EDGE><SOURCE vertex="1" offset="0 0"/><TARGET vertex="1" offset="1 0"/></EDGE> <EDGE><SOURCE vertex="1" offset="0 0"/><TARGET vertex="1" offset="-1 1"/></EDGE> </UNITCELL> I am confused about this. I don't why it is specified as this. It doesn't seem natural to me. Just from the offset, it will form a right triangle, rather than an equilateral triangle. Also, does the length of the offset matter? 2, kagome, <UNITCELL name="Kagome" dimension="2"> <VERTEX><COORDINATE>0 0</COORDINATE></VERTEX> <VERTEX><COORDINATE>0.5 0</COORDINATE></VERTEX> <VERTEX><COORDINATE>0 0.5</COORDINATE></VERTEX> <EDGE><SOURCE vertex="1"/><TARGET vertex="2"/></EDGE> <EDGE><SOURCE vertex="1"/><TARGET vertex="3"/></EDGE> <EDGE><SOURCE vertex="2"/><TARGET vertex="3"/></EDGE> <EDGE><SOURCE vertex="1"/><TARGET vertex="2" offset="-1 0"/></EDGE> <EDGE><SOURCE vertex="2"/><TARGET vertex="3" offset="1 -1"/></EDGE> <EDGE><SOURCE vertex="1"/><TARGET vertex="3" offset="0 -1"/></EDGE> </UNITCELL> Why specify the coordinates as this? Why rewrite the <EDGE> once more with offset? I cannot draw a unit cell from just looking this. Could some one explain this for me? THanks very much! Gang Chen

2 1

A couple of questions more
by Justin David Peel 15 Aug '07

15 Aug '07

Hi all, Thanks for all of the help so far. I just have a couple more questions. First of all, is there a way to set a parameter to infinity? I'm not talking about the lattice size, but for instance in the Hubbard model, I would like to try setting U to infinity or at least very large. I can't set it much larger than 17000 without the program messing up. The second question I have is a little more involved. I would like to set up an open chain lattice as defined in the standard library, but with a different 'type' for each site. Actually, what I would like is for the sites to alternate between two different types. I have read all of the docs that I could find, but haven't been able to get this to work yet. I tried changing the unit cell, among other things, but when I tried out my model, it didn't work how it should with the same U's everywhere (in the Hubbard model). If anyone can help me out with this, I would really appreciate this. Thanks again for all of your kind help, Justin

1 0

Hubbard model DMRG questions
by Justin David Peel 09 Aug '07

09 Aug '07

Hello, I have a few questions and a few suggestions. So far on my first day of using ALPS, I think that it is amazingly powerful and would like to say thank you for all of the work you all have done. I have been studying DMRG by myself making my own codes, but they are fairly simple compared to this. I have figured out the basics of the Heisenberg model with the help of the tutorials and obtained the Haldane gap, but I want to make sure that I am doing the fermionic Hubbard model correctly. Here is the parameter file I am using: LATTICE="open chain lattice" SWEEPS=4 MAXSTATES=100 { MODEL="fermion Hubbard", L=8, N_total=8, U=10 } Is this correct for a simple calculation of the ground state? Also, I was looking through the Models tutorial and noticed under the "Hamiltonian descriptions" subtopic that the SITETERM has a type. Does the type correspond to a given site that it applies to? For instance, could I set the first site to have a repulsion term of U and the second term to have a repulsion term of 2*U and so on? If not, is there a way to do this? As far as suggestions, I had to figure some things out when installing under Windows/Cygwin and thought that I could offer some modifications to the installation instructions. The biggest obstacle was setting the $PATH to include the binaries folder in the Installation folder. I can give some other suggested changes regarding if they are wanted. I also noticed that the particle in a box demo wasn't working for me. (The error message using parm1 in the tutorials folder is "EXCEPTION: bad lexical cast: source type value could not be interpreted as target" when following the instructions of the tutorial page) Lastly, I was wondering if the links could be made to work for the user workshop page if the files exist. I think they would be a great help. Thank you for your efforts and for any help you can offer, Justin Peel

3 2

Installation Problems using Cygwin in Windows
by Justin David Peel 02 Aug '07

02 Aug '07

Hello, I have been trying to install alps in cygwin, but I'm afraid that the directions don't seem to be working very well for me. I know that most people use a server install, but I would like to use ALPS in Windows. I have attached the config log installing the libraries. I installed version 1.2.2. There appears to be an error: 'XDR' does not name a type. I tried doing the rest of the install, but nothing really seemed to work. Also, the installation instructions are not very clear - at least for someone who isn't very experienced with cygwin. I had to change parts of it to get things to work at all. The biggest discrepancies are with the directory file names which don't seem to be very consistent. Anyway, I would really like to use ALPS for to do more DMRG because I have been writing my own DMRG code so far. Thank you so much for any help you can offer. Sincerely, Justin

2 1