Appion

• Changing Table Names:
  

  ApRefinementData -> ApRefineIterData #fix confusing name
  ApRefinementRunData -> ApRefineRunData #shorten name
  ApParticleClassificationData -> ApRefineParticleData #fix confusing name
  ApRefinementParamsData -> ApEmanRefineIterData  #EMAN only params
  
  ApStackParticlesData -> ApStackParticleData #rm plural
  ApAlignParticlesData -> ApAlignParticleData #rm plural
  ApClusteringParticlesData -> ApClusteringParticleData #rm plural
  

• Columns Moving to Different Table:
  

  ApRefinementParamsData.symmetry -> ApRefineIterData.symmetry #these are global params not EMAN only
  ApRefinementParamsData.mask -> ApRefineIterData.mask #these are global params not EMAN only
  ApRefinementParamsData.imask -> ApRefineIterData.imask #these are global params not EMAN only
  

• Merging Columns:
  

  ApRefinementData.classAverage
  ApRefinementData.emanClassAvg
  ApRefinementData.MsgPGoodClassAvg #barely used
  ApRefinementData.SpiCoranGoodClassAvg
  

  There are 4 columns for class averages, I want to have a more generic 2 column system of:
  

  ApRefinementData.classAverage
  ApRefinementData.classAverage2
  

  Or we can call it finalClassAverage, filteredClassAverage

ApParticleClassificationData.coran_keep -> ApRefineParticleData.filter_keep
ApParticleClassificationData.msgp_keep -> ApRefineParticleData.filter_keep
ApParticleClassificationData.thrown_out ~~ ApRefineParticleData.refine_keep #need to invert value

• Changing Column Names:
  

  ApRefinementData.refinementParams -> ApRefineIterData.emanParams #eman only table
  ApRefinementData.refinementRun -> ApRefinementData.refineRun #shorten name
  ApRefinementRunData.name -> ApRefineRunData.runname #make consistent with other tables
  

• New Calculated Columns:
  

  ApRefineRunData.num_iter #number of iterations
  ApStackData.boxsize #final boxsize of stack, makes it easier to get/change rather than calculate
  

• Project columns:
  

  *.project|projects|project -> *.project
  

  Can we do this?

what is *.project? I don't see what that accomplishes since the table and column for it is not in sinedon format.

Merging columns
ApRefinementData.classAverage
ApRefinementData.emanClassAvg
ApRefinementData.MsgPGoodClassAvg #barely used
ApRefinementData.SpiCoranGoodClassAvg

I thought the last three were divided to make it easier to query individually. I don't remember the details.
If we are sure that package name is sufficient for knowing which ones were generated from what method, generic names are indeed simpler.

Actually the sinedon format would be:

REF|project|projects|projectId

project|projects|project

REF|project|projects|DEF_id

Currently the project Table is not compatible with Sinedon and all the fields linking to it are just an integers.

Comment on merging columns by merging we get some advantages:

• Programs that make substack new volumes based on class averages only need to link to one column not 3 or 4.
• The Recon Report page/particledata.inc will have less IF statements
• If type is of concern we can create a new field, classAverage2type, that will fix this

Actually sinedon format would be `REF|projectdata|projects|xxxx` xxxx refers to the key you used in the appiondata.py
where projectdata refers to the module name in sinedon.cfg not the database name.

About merging columns, I agree that it should work fine.

MyAMI Branch with Schema changes has been created.

svn co http://emg.nysbc.org/svn/myami/branches/may2010schema/ may2010schema/

Added schema-14077.py to do this. It can be done any time, but the code is already ready written.

source:trunk/dbschema/schema-14077.py

Why can't I update this issue to closed?

You must close #345 before you can close this one because it is blocking this.