On Nov 24, 2006, at 12:28 PM, Cristian Degli Esposti Boschi wrote:
I recently installed ALPS (lib+apps), release version 1.2.2 and played a while also with a Live CD shipped with 1.3.2.
Am I right that the tutorial on the web site are referred to the pre-release 1.3 ?
Yes, the tutorial for 1.2.2 is included with the distribution.
In particular, using the same input files, with sparsediag of 1.2.2 I cannot produce any output as far as correlations are concerned while with 1.3.2 I can. What I find misleading is that with 1.2.2 I had no complain message about the parameters MEASURE_CORRELATIONS, that seem to be simply ignored.
Yes, these are additional parameters, and extra parameters that are not needed always get ignored so that the same input file can be used with many applications - parameters for other applications are ignored.
Well, apart from that, I find the implementation of TOTAL_MOMENTUM very useful (again it seems to be active only with 1.3.2).
Correct
I wonder if the eigenvalue can be selected from input...how?
This will be possible in the final version of 1.3 We are fixing a bug with selecting the momentum at the moment.
More generally, which is the syntax to select two or more compatible quantum numbers?
CONSERVED_QUANTUMNUMBERS="???";
Yes, here you specify the quantum numbers that are conserved. They need to be defined in the model description as possible conserved quantum numbers. look, e.g. at the spin model in 1.3
This will also be documented in the final version of 1.3
Best regards
Matthias Troyer