Removes all vectors from the dataset file which have an associated dual variable of zero. This is entirely optional--its purpose is only to shrink the model file after optimization, but before classification, by removing unnecessary information. Allowed options: -h [ --help ] display this help -i [ --input ] arg input model file -o [ --output ] arg output text file -s [ --small_clusters ] arg (=0) use size-16 instead of size-256 clusters? -a [ --active_clusters ] arg (=64) number of "active" clusters