Convert String to Datetime

Before we can perform time-series analysis, it helps greatly to convert the column to Python’s datatime object.

Load Libraries

import pandas as pd

Load Data

The data that we will use for this exercise is adapted from Bay Area Bike Share’s (now known as Ford GoBike) open data project to analyse and visualise ride information.

data = [{'Trip ID': 913465, 'Start Date': '9/1/2015  00:10:00', 
         'End Date': '9/1/2015  00:23:00', 
         'Start Station': 'San Francisco Caltrain 2 (330 Townsend)'},
        {'Trip ID': 913466, 'Start Date': '9/1/2015  00:15:00', 
         'End Date': '9/1/2015  00:31:00', 
         'Start Station': 'Clay at Battery'}]

df = pd.DataFrame(data)
df.dtypes

End Date         object
Start Date       object
Start Station    object
Trip ID           int64
dtype: object

By default, Pandas does not parse the strings to datetime objects. We will need to do so ourselves.

Parsing String to Datetime

Pandas’ to_datetime function transforms the input strings to datetime objects. You can find more information on how to create the format parameter on Python’s strftime() and strptime() documentation page.

df['Start Date'] = pd.to_datetime(df['Start Date'], 
                                  format='%m/%d/%Y %H:%M:%S')
df['End Date'] = pd.to_datetime(df['End Date'], 
                                format='%m/%d/%Y %H:%M:%S')

df.dtypes

End Date         datetime64[ns]
Start Date       datetime64[ns]
Start Station            object
Trip ID                   int64
dtype: object

Having the Start Date and End Date columns in datetime allows for far easier manipulation of the data.

df['Ride Time'] = df['End Date'] - df['Start Date']
df.loc[:, ['Start Date', 'End Date', 'Ride Time']].head()