When you are using df["column"] you are getting a series object and the split() attributes is no applicable to that type of object. When using df["column"].str you are actually accessing the values of the series and python is identifying them as strings, so you can preform string functions on those objects.
与恶龙缠斗过久,自身亦成为恶龙；凝视深渊过久,深渊将回以凝视…
