使用Pandas在Python中解析日期

Question

So my question is when I run this code for first time and it was giving me the results correctly ie in the format of 2013-01-23.

But when i tried running this code next time I was not getting the correct result (giving the output as 23/01/2013).

Why is it different the second time?

from pandas import *
fec1 = read_csv("/user_home/w_andalib_dvpy/sample_data/sample.csv")

def convert_date(val):
    d, m, y = val.split('/')
    return datetime(int(y),int(m),int(d))

# FECHA is the date column name in raw file. format: 23/01/2013
fec1.FECHA.map(convert_date)
fec1.FECHA

Answer 1

Parsing dates with pandas can be done at the time you read the csv by passing parse_dates=['yourdatecolumn'] and date_parser=convert_date to the pandas.read_csv method .

Doing it this way is a much faster operation than loading the data, then parsing the dates.

The reason you get different outputs when you do the same operation twice is probably due to that when you parse the dates, you take D/M/Y as input, but have Y/M/D as output. it basically flips the D and Y every time.