Getting NA while taking mean using groupby in pandas

Question

I have a DataFrame like this

_d=pd.DataFrame({\'key1\':[1,2,3,4,5,5,5,6,6],
    \'key2\':[10,20,30,40,50,51,52,60,61],
    \'key3\':[100,200,300,400,500,501,502,6