How to Convert a Pandas Timezone-Aware DateTimeIndex to a Naive Timestamp in a Specific Timezone?

Convert Pandas Timezone-Aware DateTimeIndex to Naive Timestamp in Specified Timezone

Pandas offers the tz_localize function to create a timezone-aware Timestamp or DateTimeIndex. But what if you want to undo this operation? How do you convert a timezone-aware Timestamp to a naive one without altering its timezone?

Problem:

Let's consider a timezone-aware DateTimeIndex:

<code class="python">t = pd.date_range(start="2013-05-18 12:00:00", periods=10, freq='s', tz="Europe/Brussels")</code>

Setting the timezone to None removes the timezone but converts the timestamps to UTC, altering the visible time.

Solution (Pandas 0.15.0 ):

Since Pandas 0.15.0, the tz_localize method can be used to set the timezone to None. This removes the timezone information while preserving the local time:

<code class="python">t.tz_localize(None)</code>

This results in a naive local time DateTimeIndex.

Alternatively, tz_convert(None) can be used to convert to naive UTC time:

<code class="python">t.tz_convert(None)</code>

Performance Considerations:

The tz_localize(None) method is significantly more efficient than using a loop to replace the timezone information:

<code class="python">%timeit t.tz_localize(None)
1000 loops, best of 3: 233 µs per loop

%timeit pd.DatetimeIndex([i.replace(tzinfo=None) for i in t])
10 loops, best of 3: 99.7 ms per loop</code>

The above is the detailed content of How to Convert a Pandas Timezone-Aware DateTimeIndex to a Naive Timestamp in a Specific Timezone?. For more information, please follow other related articles on the PHP Chinese website!