Find all columns of dataframe in Pandas whose type is float, or a particular type?

This is conciser:

# select the float columns
df_num = df.select_dtypes(include=[np.float])
# select non-numeric columns
df_num = df.select_dtypes(exclude=[np.number])