>>> import pytz
>>> pytz.timezone('Asia/Hong_Kong')
<DstTzInfo 'Asia/Hong_Kong' LMT+7:37:00 STD>
A seven hour and 37 minute offset? This is a little strange, does anyone experience the same issue?
In fact I'm getting different behavior between
import pytz
from datetime import datetime
hk = pytz.timezone('Asia/Hong_Kong')
dt1 = datetime(2012,1,1,tzinfo=hk)
dt2 = hk.localize(datetime(2012,1,1))
if dt1 > dt2:
print "Why?"
Time zones and offsets change over the years. The default zone name and offset delivered when pytz creates a timezone object are the earliest ones available for that zone, and sometimes they can seem kind of strange. When you use localize to attach the zone to a date, the proper zone name and offset are substituted. Simply using the datetime constructor to attach the zone to the date doesn't allow it to adjust properly.
While I'm sure historic changes in timezones are a factor, passing pytz timezone object to the DateTime constructor results in odd behavior even for timezones that have experienced no changes since their inception.
import datetime
import pytz
dt = datetime.datetime(2020, 7, 15, 0, 0, tzinfo= pytz.timezone('US/Eastern'))
produces
2020-07-15 00:00:00-04:56
Creating the datetime object then localizing it produced expected results
import datetime
import pytz
dt = datetime.datetime(2020, 7, 15, 0, 0)
dt_local = timezone('US/Eastern').localize(dt)
produces
2020-07-15 00:00:00-04:00
Coming here nearly 10 years later, I think it's worth a note that we can now exclusively utilize the Python 3.9+ standard library to handle time zones, without a "localize trap".
Use the zoneinfo module to set and replace the tzinfo however you like, ex:
from datetime import datetime
from zoneinfo import ZoneInfo
hk = ZoneInfo('Asia/Hong_Kong')
print(repr(hk))
# zoneinfo.ZoneInfo(key='Asia/Hong_Kong')
dt1 = datetime(2012,1,1,tzinfo=hk)
print(dt1)
# 2012-01-01 00:00:00+08:00
there is a deprecation shim for pytz
Alternatives, if you're not able to use zoneinfo:
for Python < 3.9, there's backports.zoneinfo
you could also use dateutil, which follows the same semantics as zoneinfo
Note for pandas users:
pandas (v1.4.1) is still using pytz internally, and seems to have some trouble with ZoneInfo timezone objects
Related
>>> import pytz
>>> pytz.timezone('Asia/Hong_Kong')
<DstTzInfo 'Asia/Hong_Kong' LMT+7:37:00 STD>
A seven hour and 37 minute offset? This is a little strange, does anyone experience the same issue?
In fact I'm getting different behavior between
import pytz
from datetime import datetime
hk = pytz.timezone('Asia/Hong_Kong')
dt1 = datetime(2012,1,1,tzinfo=hk)
dt2 = hk.localize(datetime(2012,1,1))
if dt1 > dt2:
print "Why?"
Time zones and offsets change over the years. The default zone name and offset delivered when pytz creates a timezone object are the earliest ones available for that zone, and sometimes they can seem kind of strange. When you use localize to attach the zone to a date, the proper zone name and offset are substituted. Simply using the datetime constructor to attach the zone to the date doesn't allow it to adjust properly.
While I'm sure historic changes in timezones are a factor, passing pytz timezone object to the DateTime constructor results in odd behavior even for timezones that have experienced no changes since their inception.
import datetime
import pytz
dt = datetime.datetime(2020, 7, 15, 0, 0, tzinfo= pytz.timezone('US/Eastern'))
produces
2020-07-15 00:00:00-04:56
Creating the datetime object then localizing it produced expected results
import datetime
import pytz
dt = datetime.datetime(2020, 7, 15, 0, 0)
dt_local = timezone('US/Eastern').localize(dt)
produces
2020-07-15 00:00:00-04:00
Coming here nearly 10 years later, I think it's worth a note that we can now exclusively utilize the Python 3.9+ standard library to handle time zones, without a "localize trap".
Use the zoneinfo module to set and replace the tzinfo however you like, ex:
from datetime import datetime
from zoneinfo import ZoneInfo
hk = ZoneInfo('Asia/Hong_Kong')
print(repr(hk))
# zoneinfo.ZoneInfo(key='Asia/Hong_Kong')
dt1 = datetime(2012,1,1,tzinfo=hk)
print(dt1)
# 2012-01-01 00:00:00+08:00
there is a deprecation shim for pytz
Alternatives, if you're not able to use zoneinfo:
for Python < 3.9, there's backports.zoneinfo
you could also use dateutil, which follows the same semantics as zoneinfo
Note for pandas users:
pandas (v1.4.1) is still using pytz internally, and seems to have some trouble with ZoneInfo timezone objects
>>> import pytz
>>> pytz.timezone('Asia/Hong_Kong')
<DstTzInfo 'Asia/Hong_Kong' LMT+7:37:00 STD>
A seven hour and 37 minute offset? This is a little strange, does anyone experience the same issue?
In fact I'm getting different behavior between
import pytz
from datetime import datetime
hk = pytz.timezone('Asia/Hong_Kong')
dt1 = datetime(2012,1,1,tzinfo=hk)
dt2 = hk.localize(datetime(2012,1,1))
if dt1 > dt2:
print "Why?"
Time zones and offsets change over the years. The default zone name and offset delivered when pytz creates a timezone object are the earliest ones available for that zone, and sometimes they can seem kind of strange. When you use localize to attach the zone to a date, the proper zone name and offset are substituted. Simply using the datetime constructor to attach the zone to the date doesn't allow it to adjust properly.
While I'm sure historic changes in timezones are a factor, passing pytz timezone object to the DateTime constructor results in odd behavior even for timezones that have experienced no changes since their inception.
import datetime
import pytz
dt = datetime.datetime(2020, 7, 15, 0, 0, tzinfo= pytz.timezone('US/Eastern'))
produces
2020-07-15 00:00:00-04:56
Creating the datetime object then localizing it produced expected results
import datetime
import pytz
dt = datetime.datetime(2020, 7, 15, 0, 0)
dt_local = timezone('US/Eastern').localize(dt)
produces
2020-07-15 00:00:00-04:00
Coming here nearly 10 years later, I think it's worth a note that we can now exclusively utilize the Python 3.9+ standard library to handle time zones, without a "localize trap".
Use the zoneinfo module to set and replace the tzinfo however you like, ex:
from datetime import datetime
from zoneinfo import ZoneInfo
hk = ZoneInfo('Asia/Hong_Kong')
print(repr(hk))
# zoneinfo.ZoneInfo(key='Asia/Hong_Kong')
dt1 = datetime(2012,1,1,tzinfo=hk)
print(dt1)
# 2012-01-01 00:00:00+08:00
there is a deprecation shim for pytz
Alternatives, if you're not able to use zoneinfo:
for Python < 3.9, there's backports.zoneinfo
you could also use dateutil, which follows the same semantics as zoneinfo
Note for pandas users:
pandas (v1.4.1) is still using pytz internally, and seems to have some trouble with ZoneInfo timezone objects
I have a naïve datetime in hand and I want to promote it to include the system timezone. I see lots of examples where the timezone is known a priori and the datetime is localized against it. How do I obtain the local timezone if I don't know it in advance?
local_tz = ???
dt = local_tz.localize(naive_time)
You can get the appropriate local_tz by doing:
import time
import pytz
local_tz = pytz.timezone(time.tzname[0])
EDIT: It appears that time.tzname[0] isn't always a name that corresponds to those found in pytz.all_timezones
The simplest solution is to use the tzlocal package, which uses a number of tricks to find the local timezone from the system, and can also localize for you in place of pytz:
import tzlocal
local_tz = tzlocal.get_localzone()
local_tz.localize(naive_time)
Here is as close as I've managed to get:
local_tz = dateutil.tz.tzlocal()
timestamp = timestamp.replace(tzinfo=local_tz)
Unfortunately local_tz doesn't have a localize() method so I resorted to using replace().
I'm trying to convert a timestamp with a specific timezone(Europe/Paris) to a datetime format in UTC.
From my laptop it works with the solution below but when I'm executing my code in a remote server(AWS- Lambda function in Ireland), I've a shift of 1 hour because the local timezone of the server is different from mine.
How can I have a code who can work on my laptop and at the same time in a remote server(dynamically handle local timezone)?
import pytz
import datetime
def convert_timestamp_in_datetime_utc(timestamp_received):
utc = pytz.timezone('UTC')
now_in_utc = datetime.datetime.utcnow().replace(tzinfo=utc).astimezone(pytz.UTC)
fr = pytz.timezone('Europe/Paris')
new_date = datetime.datetime.fromtimestamp(timestamp_received)
return fr.localize(new_date, is_dst=None).astimezone(pytz.UTC)
Thanks
I am not sure what timestamp_received is, but I think what you want is utcfromtimestamp()
import pytz
from datetime import datetime
def convert_timestamp_in_datetime_utc(timestamp_received):
dt_naive_utc = datetime.utcfromtimestamp(timestamp_received)
return dt_naive_utc.replace(tzinfo=pytz.utc)
For completeness, here is another way to accomplish the same thing by referencing python-dateutil's tzlocal time zone:
from dateutil import tz
from datetime import datetime
def convert_timestamp_in_datetime_utc(timestamp_received):
dt_local = datetime.fromtimestamp(timestamp_received, tz.tzlocal())
if tz.datetime_ambiguous(dt_local):
raise AmbiguousTimeError
if tz.datetime_imaginary(dt_local):
raise ImaginaryTimeError
return dt_local.astimezone(tz.tzutc())
class AmbiguousTimeError(ValueError):
pass
class ImaginaryTimeError(ValueError):
pass
(I added in the AmbiguousTimeError and ImaginaryTimeError conditions to mimic the pytz interface.) Note that I'm including this just in case you have a similar problem that needs to make reference to the local time zone for some reason - if you have something that will give you the right answer in UTC, it's best to use that and then use astimezone to get it into whatever local zone you want it in.
How it works
Since you expressed that you were still a bit confused about how this works in the comments, I thought I would clarify why this works. There are two functions that convert timestamps to datetime.datetime objects, datetime.datetime.fromtimestamp(timestamp, tz=None) and datetime.datetime.utcfromtimestamp(timestamp):
utcfromtimestamp(timestamp) will give you a naive datetime that represents the time in UTC. You can then do dt.replace(tzinfo=pytz.utc) (or any other utc implementation - datetime.timezone.utc, dateutil.tz.tzutc(), etc) to get an aware datetime and convert it to whatever time zone you want.
fromtimestamp(timestamp, tz=None), when tz is not None, will give you an aware datetime equivalent to utcfromtimestamp(timestamp).replace(tzinfo=timezone.utc).astimezone(tz). If tz is None, instead of converting too the specified time zone, it converts to your local time (equivalent to dateutil.tz.tzlocal()), and then returns a naive datetime.
Starting in Python 3.6, you can use datetime.datetime.astimezone(tz=None) on naive datetimes, and the time zone will be assumed to be system local time. So if you're developing a Python >= 3.6 application or library, you can use datetime.fromtimestamp(timestamp).astimezone(whatever_timezone) or datetime.utcfromtimestamp(timestamp).replace(tzinfo=timezone.utc).astimezone(whatever_timezone) as equivalents.
Specifically, given the timezone of my server (system time perspective) and a timezone input, how do I calculate the system time as if it were in that new timezone (regardless of daylight savings, etc)?
import datetime
current_time = datetime.datetime.now() #system time
server_timezone = "US/Eastern"
new_timezone = "US/Pacific"
current_time_in_new_timezone = ???
If you know your origin timezone and the new timezone that you want to convert it to, it turns out to be very straightforward:
Make two pytz.timezone objects, one for the current timezone and one for the new timezone e.g. pytz.timezone("US/Pacific"). You can find a list of all official timezones in pytz library: import pytz; pytz.all_timezones
Localize the datetime/timestamp of interest to the current timezone e.g.
current_timezone = pytz.timezone("US/Eastern")
localized_timestamp = current_timezone.localize(timestamp)
Convert to new timezone using .astimezone() on the newly localized datetime/timestamp from step 2 with the desired timezone's pytz object as input e.g. localized_timestamp.astimezone(new_timezone).
Done!
As a full example:
import datetime
import pytz
# a timestamp I'd like to convert
my_timestamp = datetime.datetime.now()
# create both timezone objects
old_timezone = pytz.timezone("US/Eastern")
new_timezone = pytz.timezone("US/Pacific")
# two-step process
localized_timestamp = old_timezone.localize(my_timestamp)
new_timezone_timestamp = localized_timestamp.astimezone(new_timezone)
# or alternatively, as an one-liner
new_timezone_timestamp = old_timezone.localize(my_timestamp).astimezone(new_timezone)
Bonus: but if all you need is the current time in a specific timezone, you can conveniently pass that timezone directly into datetime.now() to get the current times directly:
datetime.datetime.now(new_timezone)
When it comes to needing timezones conversions generally, I would strongly advise that one should store all timestamps in your database in UTC, which has no daylight savings time (DST) transition. And as a good practice, one should always choose to enable time zone support (even if your users are all in a single time zone!). This will help you avoid the DST transition problem that plagues so much software today.
Beyond DST, time in software can be generally quite tricky. To get a sense of just how difficult it is to deal with time in software in general, here is a potentially enlightening resource: http://yourcalendricalfallacyis.com
Even a seemingly simple operation as converting a datetime/timestamp into a date can become non-obvious. As this helpful documentation points out:
A datetime represents a point in time. It’s absolute: it doesn’t depend on anything. On the contrary, a date is a calendaring concept. It’s a period of time whose bounds depend on the time zone in which the date is considered. As you can see, these two concepts are fundamentally different.
Understanding this difference is a key step towards avoiding time-based bugs. Good luck.
With Python 3.9, the standard lib has all you need: zoneinfo. pytz is not needed anymore (deprecated; -> pytz deprecation shim).
Ex:
from datetime import datetime
from zoneinfo import ZoneInfo
server_timezone = "US/Eastern"
new_timezone = "US/Pacific"
current_time = datetime.now(ZoneInfo(server_timezone))
# current_time_in_new_timezone = ???
current_time_in_new_timezone = current_time.astimezone(ZoneInfo(new_timezone))
That gives you for example
print(current_time.isoformat(timespec='seconds'))
# 2021-10-04T02:42:54-04:00
print(repr(current_time))
# datetime.datetime(2021, 10, 4, 2, 42, 54, 40600, tzinfo=zoneinfo.ZoneInfo(key='US/Eastern'))
print(current_time_in_new_timezone.isoformat(timespec='seconds'))
# 2021-10-03T23:42:54-07:00
print(repr(current_time_in_new_timezone))
# datetime.datetime(2021, 10, 3, 23, 42, 54, 40600, tzinfo=zoneinfo.ZoneInfo(key='US/Pacific'))
How do you convert datetime/timestamp from one timezone to another timezone?
There are two steps:
Create an aware datetime objects from the system time and timezone e.g., to get the current system time in the given timezone:
#!/usr/bin/env python
from datetime import datetime
import pytz
server_timezone = pytz.timezone("US/Eastern")
server_time = datetime.now(server_timezone) # you could pass *tz* directly
Note: datetime.now(server_timezone) works even during ambiguous times e.g., during DST transitions while server_timezone.localize(datetime.now()) may fail (50% chance).
If you are sure that your input time exists in the server's timezone and it is unique then you could pass is_dst=None to assert that:
server_time = server_timezone.localize(naive_time, is_dst=None)
It raises an exception for invalid times.
If it is acceptable to ignore upto a day error (though typically an error due to DST is around an hour) then you could drop is_dst parameter:
server_time = server_timezone.normalize(server_timezone.localize(naive_time))
.normalize() is called to adjust non-existing times (local time in the gap, during "spring forward" transitions). If the time zone rules haven't changed; your server shouldn't generate non-existing times. See "Can I just always set is_dst=True?"
Convert an aware datetime object to the target timezone tz:
tz = pytz.timezone("US/Pacific")
server_time_in_new_timezone = server_time.astimezone(tz)
In case you want to convert a timestamp from one timezone to another, you can use this code:
from datetime import datetime
from zoneinfo import ZoneInfo
from_timezone = ZoneInfo('Europe/Moscow') # UTC-3
to_timezone = ZoneInfo('Asia/Tbilisi') # UTC-4
dt = datetime.fromtimestamp(timestamp, to_timezone)
result_timestamp = int(dt.replace(tzinfo=from_timezone).timestamp())
For example, if you take timestamp = 529635600 (1986-14-10 04:00:00 in Moscow) and run this code, you will get result_timestamp = 529639200 (1986-14-10 05:00:00 in Tbilisi).