[Wien] the results about electron density difference

[yxl@email.jlu.edu.cn]

Dear users,
   I have asked the question last month, but nobody responded.
I am studying the electron density difference, from where we can know some about the bonding. For example SnO2, the bonding orbital should be Sn:4s 4p, O:2p. But I do not think my following results are reasonable.
  When I set the Emin in case.in2 to 0.01(Sn:4s 4p, O:2p), the electron density difference located on Sn and O atoms are all minus, and the maxmum is -15.
  When I set the Emin to -1.0(Sn:4s 4p, O:2s 2p), the electron density difference located on Sn are all positive,the center of O are still minus,but little positive density difference appear around the O atoms. The value is from -1.21~0.5.
So I have two questions:
1.When the Emin became form 0.01 to -1.0, the density should much more and include the density appears in the plot of 0.01. Why the minus density difference located on Sn disappeared, and why the value became to -1.2 from -14?
2.The positive density difference on O means that O atoms should get electrons. I think the added electrons should be in O 2p. Why there was no positive density difference on O when Emin was set to 0.01, but appeared when the Emin was set to -1.0?   
3.Why the electrons increased for Sn, decreased for O?
Any suggestions will be appreciated.
yours sincerely,
hongxia