temporarily use first dimension for calculating E field location

This is correct for node centered electric fields where all components
are located at the same place in a cell, but wrong for edge electric
fields.

src/particle_updaters/electrostatic.jl

@@ -56,7 +56,7 @@ function step!(step::ElectrostaticParticlePush)
         )
 
         for I in Is
-            grid_cell_coord = cell_index_to_cell_coords(step.E, I)
+            grid_cell_coord = cell_index_to_cell_coords(step.E, I, 1)
             dist = Tuple(particle_cell_coord .- grid_cell_coord)
             interp_weights = step.interpolation_function.(dist)
             interp_weight = prod(interp_weights)