http://membrane.urmc.rochester.edu/Software/WHAM/WHAM.html I have written a fast, memory efficient implementation of the Weighted Histogram Analysis Method (WHAM). I've compiled and tested on a number of platforms and it's vanilla ANSI C, so the odds are good it'll work almost anywhere (or at least anywhere sufficiently unix-like). It is released under the GPL and BSD licenses, so you can take your pick of whichever you're more comfortable with. However, if you're using the code, I'd appreciate it if you'd let me know. Suggestions, patches, and complaints are welcome (alan_grossfield [AT] urmc.rochester.edu). I wrote this code for my own purposes, but several groups now use it, and I think it's pretty solid.<br/>The current version of the source is available here. A fairly detailed manual is also included (pdf also available). (Note: if you download the docs, you'll need to rerun latex2html to get the html docs correct; the pdf will be present out of the box). If you still have questions after reading it, you can contact me (alan_grossfield [AT] urmc.rochester.edu) and I'll do my best to help.<br/>For a simple discussion of the theory and practice of using WHAM to compute PMFs, you are welcome to download the notes from a talk I gave. iWeb 2.0.4 http://membrane.urmc.rochester.edu/Software/WHAM/WHAM_files/Bilayer.jpg http://membrane.urmc.rochester.edu/Software/WHAM/WHAM.html http://membrane.urmc.rochester.edu/Software/WHAM/Entries/2008/4/14_Version_2.0.2.html 536bd85d-ced6-4bec-be2c-278ab94f393e Mon, 14 Apr 2008 09:51:53 -0400 <a href="http://membrane.urmc.rochester.edu/wham/wham-release-2.0.2.tgz">Download tarball</a><br/>This is a bug-fix release to version 2.0.1, because the previous patch didn't actually fix the buffer overflow. This time for sure... As usual, let me know if you have any trouble. http://membrane.urmc.rochester.edu/Software/WHAM/Entries/2008/2/8_Version_2.0.1.html 3e60d45e-8ef3-4fa3-b57b-acc72cdf9b8e Fri, 8 Feb 2008 10:00:46 -0500 This is a bug-fix release to version 2.0. Two changes have been made:<br/>Output the intermediate F values when dumping the intermediate pmf. As a result, if you're not sure how far you need to go to converge, you can set the tolerance to 0 and monitor the intermediate pmfs until they converge. You could do this before, but you'd never get the final F values, which can be useful for some analysis. Requested by Chris Neale.<br/>Fixed an overflow bug in wham-2d, where I wrote one spot past the end of cum. Spotted and diagnosed by Chris Neale.<br/><a href="http://membrane.urmc.rochester.edu/wham/wham-release-2.0.1.tgz">Download version 2.0.1</a> http://membrane.urmc.rochester.edu/Software/WHAM/Entries/2007/11/1_Entry_1.html 93f62247-b8f3-4b85-abba-51b9436f946c Thu, 1 Nov 2007 10:02:36 -0400 The first new version of the code in several years has been released. Now that I've left IBM Research, I can distribute new versions of the code without the risk of polluting the license. The new version contains several updates:<br/>Significantly reworked the Monte Carlo bootstrap error estimation. Primarily, added the ability to specify the correlation time for individual time series, which makes the method more quantitatively accurate (previous versions significantly underestimated the uncertainty). Also fixed an off-by-one error (thanks to Michael Shirts for pointing it out).<br/>Improved documentation, especially with respect to use with replica exchange calculations.<br/>Made some error messages less cryptic.<br/><a href="http://membrane.urmc.rochester.edu/wham/wham-release-2.0.tgz">Download version 2.0</a>