aboutsummaryrefslogtreecommitdiff
path: root/doc/man
diff options
context:
space:
mode:
authorThomas White <taw@physics.org>2014-03-11 16:43:06 +0100
committerThomas White <taw@physics.org>2014-03-11 16:43:06 +0100
commit9df6e29846ad8d7d69069400bd3419fa4749deb6 (patch)
treecddc1627899472fac90eb1b9bb42d2841f17e883 /doc/man
parent22678ec74ccaf34c4b88b4b7ba23869b3074d55a (diff)
ambigator: Add --really-random
Diffstat (limited to 'doc/man')
-rw-r--r--doc/man/ambigator.14
1 files changed, 4 insertions, 0 deletions
diff --git a/doc/man/ambigator.1 b/doc/man/ambigator.1
index b0978fb7..89dea842 100644
--- a/doc/man/ambigator.1
+++ b/doc/man/ambigator.1
@@ -79,6 +79,10 @@ Write f and g values to \fIfilename\fR, one line per crystal, repeating all crys
.IP \fB--ncorr=\fR\fIn\fR
Use \fIn\fR correlations per crystal. The default is to correlate against every crystal. If the CC calculation is too slow, try \fB--ncorr=1000\fR. Note that this option sets the maximum number of correlations, and some crystals might not have enough common reflections to correlate to the number requested. The mean number of actual correlations per crystal will be output by the program after the CC calculation, and if this number is much smaller than \fIn\fR then this option will not have a significant effect.
+.PD 0
+.IP \fB--really-random\fR
+Be non-deterministic by seeding the random number generator (used to make the initial indexing assignments and select patterns to correlate against) from /dev/urandom. Otherwise, with single-threaded operation (\fB-j 1\fR) on the same data, the results from this program should be the same if it is re-run. Using more than one thread already introduces some non-deterministic behaviour.
+
.SH AUTHOR
This page was written by Thomas White.