are still real!).
* assembleCompleteElectrodeModel_ is more or less copied from its
  real-values counterpart, with a few required type changes
* potentialsCEM_ is resized to hold the imaginary parts in calculateK_
* imaginary parts are then stored in potentialsCEM using the "new"
  inline function TmpToImagHACK
* Warning: This commit still contains debug output

	* cast 'start' and 'end' parameters to a numpy array to allow
	  for array-based operations
	* Slightly rework the docstring

real-valued mode even if a complex modeling is requested

…lse, h2=False'

contact resistances for CEM electrodes:
	* add the setter function setContactResistances so we can
	  directly set electrode contact resistances from the Python
	  side
	* adding this function required slight changes in the internal
	  handling logic. Previously the code needed only to look for
	  contact impedances registered using the setter function, and thus it
	  was easier to decide which resistance type to actually use
          (contact impedances and contact resistances can be converted to each
          other if the electrode surface area is known). Now we solve this by an
          explicit parameter `use_cimp` which decides if the provided impedance
          (=true) or resistance (=false) vectors shall be used further.
	* setter functions are overwritten by input from files
	  contactImpedance.map and contactResistance.map.
	* resistance input always overwrites impedance input

…esistances parameters that can be used to set CEM contact impedances and contact resistances

if-statement and make output more consistent

contact impedances and resistances with the real-valued version of this
function

and setting of multiple columns:

	data = pg.DataContainer()
	data[['a', 'b']] = [[1, 2, 3], [4, 5, 6]]
	print(data[['a', 'b']])

Also add a few tests for this new functionality

modeling using Pygimli