# Timezone of datasets

**URL:** <https://community.drivendata.org/t/timezone-of-datasets/450>\
**Category:** From Fog Nets to Neural Nets\
**Created:** [March 6, 2016, 5:25pm UTC](https://community.drivendata.org/t/timezone-of-datasets/450 "2016-03-06T17:25:40Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![oliviers](https://avatars.discourse-cdn.com/v4/letter/o/f07891/32.png) [@oliviers](https://community.drivendata.org/u/oliviers)\
**Post date:** [March 6, 2016, 5:25pm UTC](https://community.drivendata.org/t/timezone-of-datasets/450/1 "2016-03-06T17:25:40Z")

</div>

Hello,

Can you confirm that the timestamp of _all_ the datasets are in the same timezone?

---

<div class="post-metadata">

**Author:** ![bull](https://yyz2.discourse-cdn.com/flex028/user_avatar/community.drivendata.org/bull/32/8_2.png) [@bull](https://community.drivendata.org/u/bull)\
**Post date:** [March 6, 2016, 5:34pm UTC](https://community.drivendata.org/t/timezone-of-datasets/450/2 "2016-03-06T17:34:44Z")

</div>

Hi @oliviers, good question. According to the details we have, all of the datasets _should_ be in local time, and there is only one time zone in Morocco. Let us know if you find any indications that this is not the case.

---

<div class="post-metadata">

**Author:** ![oliviers](https://avatars.discourse-cdn.com/v4/letter/o/f07891/32.png) [@oliviers](https://community.drivendata.org/u/oliviers)\
**Post date:** [March 29, 2016, 4:43am UTC](https://community.drivendata.org/t/timezone-of-datasets/450/3 "2016-03-29T04:43:11Z")

</div>

The description state  
"Date / Local time - Local time in this location. Summer time (Daylight Saving Time) is taken into consideration"

This is unlikely, for 2 reasons:  
a/ this forbids to retrieve the “real” time  
b/ if it was the case, 2:00 , 2:30 should not exst the day STD =\> DST, and they should appear twice on DST =\> STD

Morocco is especially trikcy regarding DST because, ramadam falling in summer in 2014 + 2015, DST was _reversed_ ( probably to make the fast more bearable ), so there should be 2 “holes” 2:00-300 and 2 “duplicates” 2:00-3:00 each summer this year.

It probably does not make a big difference on the forecasts, but for the sake of it it would be nice to know how the data from the “Tme” column is build for each file. Maybe it is just random and guessing is part of the challenge?

---

<div class="post-metadata">

**Author:** ![bull](https://yyz2.discourse-cdn.com/flex028/user_avatar/community.drivendata.org/bull/32/8_2.png) [@bull](https://community.drivendata.org/u/bull)\
**Post date:** [March 29, 2016, 4:58pm UTC](https://community.drivendata.org/t/timezone-of-datasets/450/4 "2016-03-29T16:58:08Z")

</div>

Good question–unfortunately, we don’t know _for sure_ what happened during the time changes, so we have to ask the data. There are two datasets that have enough granularity to look at these time changes–the fog net data, and the weather data from Agadir.

From looking at the data, it looks like both datasets actually ignore the Ramadan time changes in 2014/2015. Neither one has any anomalies on the [dates](http://www.timeanddate.com/time/change/morocco/casablanca?year=2014) we expect to see [changes](http://www.timeanddate.com/time/change/morocco/casablanca?year=2015).

The only time that I spot a DST/STD change 2015-03-29, where the fog net data leaves out 02:00 - 03:00 and the Agadir data leaves out 03:00 - 04:00.

Would be happy to hear other theories if you have them!