Use correct de broglie thermal wavelength in density script

These changes were discussed in: https://gitlabph.physik.fu-berlin.de/ag-netz/mdtools/issues/19.

The implementation might not be perfect yet. As @aschlaich suggested I used the particle masses from the topology. But I am not a 100% sure how the script should behave when multiple groups are given. The current approach calculates the mass only when there is only one group and sets the mass to NaN otherwise (which results in a NaN chem potential).

Note that I also changed two lines in the _calculate_results method. The density script crashed when -zpos was given.

For a test trajectory I get the expected offset of 36.5 KJ/mol