TypeError: strptime() argument 1 must be string, not float

Question

Good morning!

I have a series of events with associated dates. The event dates are stored as a series of string values in a column within a dataframe that I have loaded into Python. The dataframe contains other columns with values. I would like to convert the values in column "event" into datetime objects, store them in a list, and then plot that list with matplotlib to create a timeseries.

My dataframe looks like this:

         date         value_1    event        other_event
37       07/02/2015   265.09     07/02/2015   NaN
38       08/02/2015   278.59     08/02/2015   NaN
156      06/06/2015   146.07     06/06/2015   NaN
180      30/06/2015   133.56     30/06/2015   NaN
243      01/09/2015   280.27     01/09/2015   01/09/2015

Python tells me that the column data is Name: event, dtype: object, which I assume means it contains string values. I also have the line df.event.apply(str) in my code, which I think will convert the values in my event column into string values.

Then I have this code:

FMT = '%d/%m/%Y'
event_list = []

for i in range(0, len(event)):
    event_list.append(datetime.datetime.strptime(event[i], FMT)) 

However, this line returns an error:

Traceback (most recent call last):

  File "<ipython-input-39-e778a465e858>", line 2, in <module>
    event_list.append(datetime.datetime.strptime(event[i], FMT))

TypeError: strptime() argument 1 must be string, not float

Any advice as to where I'm going wrong will be gratefully received.

Answer 1

To plot the dataframe in question using matplotlib, you may convert the column in question to datetime using pandas.to_datetime first.

u = u"""i         date         value_1    event        other_event
37       07/02/2015   265.09     07/02/2015   NaN
38       08/02/2015   278.59     08/02/2015   NaN
156      06/06/2015   146.07     06/06/2015   NaN
180      30/06/2015   133.56     30/06/2015   NaN
243      01/09/2015   280.27     01/09/2015   01/09/2015"""

import io
import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv(io.StringIO(u), delim_whitespace=True)
df["event"] = pd.to_datetime(df["event"], format="%d/%m/%Y")

plt.plot(df["event"], df["value_1"])
plt.gcf().autofmt_xdate()
plt.show()