Python — Time Series Data with Pandas

Numeric, categorical and time series data are the types of data that we commonly dealt with as part of exploratory data analysis. In this article I will go through some basic operations with time series data.

The notebook for this article can be found here.

Generating Time Series Data

# Generate data with time
ts = pd.Series(np.random.rand(100), index=pd.date_range(datetime.now(), periods=100))
ts
2020-11-30 08:53:58.271878    0.566087
2020-12-01 08:53:58.271878    0.906584
2020-12-02 08:53:58.271878    0.512919
2020-12-03 08:53:58.271878    0.878789
2020-12-04 08:53:58.271878    0.942902
                                ...   
2021-03-05 08:53:58.271878    0.983846
2021-03-06 08:53:58.271878    0.289516
2021-03-07 08:53:58.271878    0.840058
2021-03-08 08:53:58.271878    0.519680
2021-03-09 08:53:58.271878    0.506116
Freq: D, Length: 100, dtype: float64

To remove the time, set normalize to True.

# Generate data without time (normalize = True)
ts = pd.Series(np.random.rand(100), index=pd.date_range(datetime.now(), periods=100, normalize=True))
ts
2020-11-30    0.403108
2020-12-01    0.386055
2020-12-02    0.904074
2020-12-03    0.705386
2020-12-04    0.527159
                ...   
2021-03-05    0.271150
2021-03-06    0.851113
2021-03-07    0.778842
2021-03-08    0.380553
2021-03-09    0.552797
Freq: D, Length: 100, dtype…