将 pandas 数据帧/系列应用于 netaddr EUI 构造函数

Question

I have a dataframe with Mac addresses, which needs converting. I first do the following to clear out any nan 's:

df = df[pd.notnull(df["MacAddr"])]["MacAddr"] # avoid null values being passed to netaddr

...then I'd like to parse into the EUI constructor like so:

from netaddr import EUI, mac_bare

mac = EUI(df)
mac.dialect = mac_bare

But this doesn't work, because EUI expects a string, not a pandas Series. Is there a way to circumvent this?

As an error I get this:

Name: MacAddr, Length: 26953, dtype: object is not str() or unicode()!

Answer 1

从系列中创建一个列表，然后将元素连接成一个字符串。

 ','.join(df[pd.notnull(df["MacAddr"])]["MacAddr"].to_list())

Answer 2

Since parsing dataframes/series isn't currently possible for the EUI constructor, I created a function for df.apply() , like so:

def main(self):
    def _eui_mac_bare_func(x):
        mac = EUI(x)
        mac.dialect = mac_bare
        return mac

series = df[pd.notnull(df["MacAddr"])]["MacAddr"] # avoid null values being passed to netaddr
new_series = series.apply(_eui_mac_bare_func) # apply function to every row
df = df.join(new_series.rename("Norm_MacAddr")) # join df and series on corresponding indices