pd.read_csv-纪元日期转换

Question

I have csv with date columns, Bill Date for example, stored as the Excel serial number for the date. So 12/15/2017 is represented in the csv as 43084. I need to convert this in my DataFrame to an actual date Pandas / Python recognizes. So, my question: Is there a more efficient method than what I've used to convert the dates?

df:
    Bill Date
0       43084

Excel stores dates as days since Jan 1, 1900 and pandas defaults to Jan 1, 1970.

diff = pd.Timestamp('1970-01-01') - pd.Timestamp('1900-01-01')
pd.to_datetime([43084 - (diff.days + 2)], unit='d')

DatetimeIndex(['2017-12-15'], dtype='datetime64[ns]', freq=None)

So, am I missing something obvious here (eg, an option for a different origin)?

Answer 1

Just do this:

 import pandas as pd
 from datetime import datetime

 df = pd.read_csv("yourdataframe.csv")
 df["Bill_Date"] = df["Bill_Date"].map(lambda x: datetime.fromordinal(datetime(1900, 1, 1).toordinal() + x - 2))
 df["Bill_Date"] = df["Bill_Date"].map(lambda x: x.strftime("%Y-%m-%d"))

 print df.head()