Appion

r15680 adds the bulk of the new files that will be needed. They are not complete.

Left todo on this task:

1. Write stackPrepForm.inc and add "binning" to the stack prep params.
2. complete the code to import parameters from prior recon runs.
3. Create a form to input the required cluster information.
4. Add the ability to select a recon stack for the Frealign method.
5. Debug docpop issues.
6. Complete work on runPreparedRecon.php.
7. Make sure we get all the stackprep params and other params that are floating around that are needed in the command.
8. Add the multi-model refinement interface. (create a seperate issue for this)

and any other todo's listed in the code...

r15683 takes care of 1 in the todo list.

r15685 adds support for multi-model refinements. - checked in without commit comment...opps.

TODO: need to add code to the menu to display completed runs, etc.

Amber,

Please change the following eman option names to its proper eman parameter name without abbreviation in its run form so that I can use them consistently in the python file in sending it to eman.

clskeep -> classkeep
clsiter -> classiter
phscls -> phasecls

--modelnames should not be the original model file name but that in ApRefineInitialModelData
--stackname should be that in ApRefineStackData

And please pass boxsize of the refine stack (the boxsize from the stack it is preped from * binning during the prep) as --boxsize=# and apix as --apix (also need to take into account of the binning)

When innerMaskRadius and outerMaskRadius is passed, they also need to include the binning (i.e., divided by the binning). The job run on the cluster should not care about the original stack/model dimension but only what is prepared. It happens often that people use the same stack but different binning to run first a test refinement at higher binning and then redo refinement in another run without binning, It should be made possible for the unbined run to import the test refinement parameters.