Bug 331894

Comment 1 Jan-Frederik Wigger 2010-12-06 08:32:20 EST

This bug is already mentioned in bug 317079.

It seems that there is a special handling for multiple populations in a "multi-population disease model". 

See
org.eclipse.stem.diseasemodels.multipopulation.provider(MultipopulationPopulationEnumeratorAdapter.java:23) 

public String[] getPopulationIdentifiers() {

MultiPopulationSIDiseaseModel dm = (MultiPopulationSIDiseaseModel)getTarget();
StringValueList list = dm.getPopulationGroups();
EList<org.eclipse.stem.core.common.StringValue> vals = list.getValues();
		
String [] res = new String[vals.size()+1];
res[0] = dm.getPopulationIdentifier();
for(int i=1;i<vals.size()+1;++i)
  res[i] = vals.get(i-1).getValue();
		
return res;
}


The method getPopulationIdentifiers() manually adds the name of the "disease model population" to the array, afterwards adds all population groups of the disease model. 

This possibly creates double entries, as taken into account in
org.eclipse.stem.util.loggers.views(NewCSVLogWriter.java:256 and :463) 

PopulationEnumeratorAdapter en = (PopulationEnumeratorAdapter)PopulationEnumeratorAdapterFactory.INSTANCE.adapt(dm, PopulationEnumerator.class);
if(en != null) {
  en.setTarget(dm);
  String [] pops = en.getPopulationIdentifiers();
  // Skip first one, it's already been added.
  if(pops.length > 1)
    for(int i=1;i<pops.length;++i) 
      populationIdentifiers.add(pops[i]);
}


The first entry is skipped, "it's already been added" (text in comment).

I cannot overlook whether it is still useful/necessary to add a disease model population that is possibly not part of the population groups (if that was the reason for this handling). Suggestions for possible solutions:

---------------------------------------------------------
1.) Avoid double entries in PopulationIdentifiers array:

org.eclipse.stem.diseasemodels.multipopulation.provider(MultipopulationPopulationEnumeratorAdapter.java:23 )

public String[] getPopulationIdentifiers() {
MultiPopulationSIDiseaseModel dm = (MultiPopulationSIDiseaseModel)getTarget();
StringValueList list = dm.getPopulationGroups();
EList<org.eclipse.stem.core.common.StringValue> vals = list.getValues();	
		
java.util.List<String> popIDs = new java.util.ArrayList<String>();
popIDs.add(dm.getPopulationIdentifier());
for(int i=0;i<vals.size();++i)
  if (!popIDs.contains(vals.get(i).getValue()))
    popIDs.add(vals.get(i).getValue());
		
String [] res = new String[popIDs.size()];
for(int i=res.length-1;i>=0;i--)
  res[i] = popIDs.get(i);
		
return res;
}

Because the LogWriter skips the first entry so far:
Delete lines org.eclipse.stem.util.loggers.views(NewCSVLogWriter.java:260-263), change :466 to 
for(int j=1;j<subpops1.length;++j) (?not sure)

------------------------------------------------------

2.)Abandon the special handling completely, replace:
org.eclipse.stem.diseasemodels.multipopulation.provider(MultipopulationPopulationEnumeratorAdapter.java:29) 

String [] res = new String[vals.size()];
for(int i=0;i<vals.size();i++)
 res[i] = vals.get(i).getValue();



Delete lines org.eclipse.stem.util.loggers.views(NewCSVLogWriter.java:260-263), change :466 to 
for(int j=1;j<subpops1.length;++j) (?not sure)

------------------------------------------------------

3.) Keep this special handling, but just do not show in the selection list:
 
org.eclipse.stem.ui.widgets(GeoViewOptionsBar.java:384) 
might need the same check and handling as described above ("Skip first one, it's already been added"). 

if(pea != null) {
pea.setTarget(decorator);
String [] pis = pea.getPopulationIdentifiers();
for (int i=(pis.length > 1)?1:0;i<pis.length;++i)
  popIds.add(pis[i]);
} 


------------------------------------------------------


Problems with 1 and 2: Maybe I do not know all effects that a change of the return array of getPopulationIdentifiers() takes on STEM. 


Jan-Frederik Wigger

Comment 2 Jan-Frederik Wigger 2010-12-06 08:33:05 EST

Created attachment 184596 [details]
screenshot of the drop down menu with the populations

Comment 3 James Kaufman 2010-12-07 14:32:37 EST

Stefan, Is this the bug you fixed yesterday?

Comment 4 Stefan Edlund 2010-12-17 16:55:02 EST

Jan-Frederik, I pretty much followed your first suggestion and that seems to work. Let me know if there's still any problems.

Thanks!

Comment 5 James Kaufman 2011-07-18 21:43:42 EDT

Complete