16.3: Extracting components of date and time

If you have a Pandas DataFrame with a column of `np.datetime64` type, you can easily extract various parts of the date, like the year, month, day, and hour.

To do this, first ensure that the column is in datetime format and, if not, use `pd.to_datetime()` to convert it. Once that's done, you can use the `dt` accessor followed by the attribute you're interested in.

For example, if your DataFrame is named `df` and the datetime column is named `timestamp`, you can access the year by using `df['timestamp'].dt.year`, the month with `df['timestamp'].dt.month`, the day with `df['timestamp'].dt.day`, and the hour with `df['timestamp'].dt.hour`. These commands will return a new Pandas Series containing just the year, month, day, or hour, which you can then use for further analysis or operations.

Here is an example.  First, let's define a DataFrame called `df`:

import pandas as pd
import numpy as np
 
# Create a sample DataFrame with np.datetime64 data
df = pd.DataFrame({
    'timestamp': pd.date_range(start='2021-01-01', end='2021-01-05', freq='12H'),
    'value': np.random.random(9)
})
 
In this case, freq = '12H' means that the frequency is every 12 hours.  The DataFrame looks like:

Let's add some new columns that contain the year, month, day, and hour of the timestamp:

# Extract year, month, day, and hour
df['year'] = df['timestamp'].dt.year
df['month'] = df['timestamp'].dt.month
df['day'] = df['timestamp'].dt.day
df['hour'] = df['timestamp'].dt.hour

Now the DataFrame looks like:

You can also use this syntax to extract rows whose dates or times satisfy some constraint.  For example, if you want to extract the rows for which hour = 12, you could use this:

filtered_df = df[df['timestamp'].dt.hour == 12]