usage: prog <particle stack>
	Generate initial model from subtomogram particles using stochastic gradient descent.

Option Type Description
--version None show program's version number and exit
--path str path
--res float target resolution
--niter int iterations
--shrink int shrink
--parallel, -P str Run in parallel, specify type:<option>=<value>:<option>=<value>. See http://blake.bcm.edu/emanwiki/EMAN2/Parallel
--ncls int number of classes
--batch int batch size
--keep float Fraction of particles to keep. will actually align more particles and use the number of particles specified by batch
--learnrate float learning rate, default 0.2
--ref str Reference volume, not required. default=none
--sym str symmetry. Only c1 unless --ref used
--classify None classify particles to the best class. there is the risk that some classes may end up with no particle. by default each class will include the best batch particles, and different classes can overlap.
--curve None Mode for filament structure refinement.
--verbose, -v int verbose level [0-9], higner number means higher level of verboseness
--ppid int ppid