[SOLVED] Can't use pendulum to parse dates in Series, but works one by one

Can't use pendulum to parse dates in Series, but works one by one

I'm trying to parse dates using pendulum. I have a TimeStamp date, so I did the following:

df['aux']=df['Date'].dt.date
df['p_date']=df.aux.apply(lambda x: pendulum.parse(x))

Which brings the following error:

AttributeError: 'DateTime' object has no attribute 'nanosecond'

But if I do, something like:

pendulum.parse(df.aux[0])

It gets parsed no problem. I thought apply(lambda x:) applied the same function to all rows of the Series , but now it isn't working. What's happening?

Sample code:

dates=pd.Series(['2018-03-20','2019-03-21'])
dates.apply(lambda x: pendulum.parse(x)) #Doesn't work
pendulum.parse(dates[0]) #Works

Solution

I think you have to use .naive() at the end

dates.apply(lambda x: pendulum.parse(x).naive()) #works

See this thread: https://github.com/sdispater/pendulum/issues/246

It seems that pandas tries to convert the timezone aware datetime to it's own timestamp representation, but that conversion isn't triggered with the naive datetime. I don't think anyone is at fault here, as a user of both pendulum and pandas it makes it difficult to use them together.