Skip non-numeric columns when aggregating MicroDataFrames

[MaxGhenis]

We should follow pandas here:

In [1]: import pandas as pd
 
In [2]: import microdf as mdf                   
                                                                                     
In [3]: d = pd.DataFrame({"x": ["a", "b"], "y": [1, 2]})          
                                               
In [4]: d.mean()
Out[4]:           
y     1.5                                        
dtype: object                                           
       
In [5]: mdf.MicroDataFrame(d).mean() 
...
TypeError: No loop matching the specified signature and casting was found for ufunc add

[MaxGhenis]

pandas raises a TypeError when running mean on a string Series.