CODEWITHSUNDEEP
pythonpandas
user3205999
user3205999

Reputation: 187

Filter pandas DataFrame by column time value

I have a pandas DataFrame with a 'date' column, which uses this format:

2015-01-01 04:00:00
2015-01-01 05:00:00
2015-01-01 06:00:00
2015-01-01 07:00:00
...
2015-01-02 04:00:00
2015-01-02 05:00:00
2015-01-02 06:00:00
2015-01-02 07:00:00

I want to filter the DataFrame so I only keep the rows with a stated time, e.g. 06:00:00

2015-01-01 06:00:00
2015-01-02 06:00:00

I tried things like

df['date'] = pd.to_datetime(df['date'])
df = df[df['date'].time() == datetime.time(6)]

But they don't work.

How can I achieve that?

Upvotes: 2

Views: 2422

Answers (1)

EdChum
EdChum

Reputation: 393903

You can use the datetime attribute accessor to access the hour attribute to filter the df once the dtype is a datetime:

In [141]:
t="""2015-01-01 04:00:00
2015-01-01 05:00:00
2015-01-01 06:00:00
2015-01-01 07:00:00
2015-01-02 04:00:00
2015-01-02 05:00:00
2015-01-02 06:00:00
2015-01-02 07:00:00"""
s = pd.read_csv(io.StringIO(t), parse_dates=[0], header=None, names=['date'])
s[s['date'].dt.hour == 6]

Out[141]:
                 date
2 2015-01-01 06:00:00
6 2015-01-02 06:00:00

Upvotes: 1

Related Questions