[Qgis-user] Query cross-table "similar" time stamps?

Sat May 18 11:17:07 PDT 2019

Hi Harrisou,

this looks a lot what I imagined (in the sense of some wicked nested
expression stuff i never would get to myself).

Am 18.05.19 um 11:42 schrieb DelazJ:
> Hi Bernd,
> Hum.. interesting. Assuming you have two layers: *tracks (id_track,
> date_track)* and *birds(id_bird, date_bird)*, the idea is to get the
> id of the highest among the date_track(s) lower than the date_bird for
> each bird. Am I right?

Well, erm, yes.

> I gave it a shot and i think i have some working expression. Here I
> display the matching track id in birds labels
>
> attribute(
>   get_feature ( 'tracks',
>                       'date_track',
>                       aggregate( layer:='tracks',
>                                       aggregate:='max',
> expression:="date_track",
>                                       filter:="date_track"<
> attribute(@parent, 'date_bird')
>                                   )
>                  ),
>   'id_track'
> )
>
> I used temporary layers with datetime field.
> Now, without sample of dataset to actually see the date formatting
> issue you were afraid of...

The layers derive from ordinary gpx-files from the phone generated with
Locus.

I labeled the point and the track points with the date format in the
screenshot https://i.stack.imgur.com/MqPhK.jpg QGIS sais the Type is
*QDateTime* and the Typename is *DateTime*. Indeed, date formats are a
pain in the arse and its no fun fumbling around with them most of the time.

Will give it a try tomorrow evening. Have to work all weekend in the
field, so already got as sunburn and tomorrow I will get washed away.

Thanx a lot so far, I will report back!

Cheers,

Bernd

>
> Hope that helps,
> Harrissou
>
> Le ven. 17 mai 2019 à 20:34, Bernd Vogelgesang
> <bernd.vogelgesang at gmx.de <mailto:bernd.vogelgesang at gmx.de>> a écrit :
>
>     Hi Mike,
>
>     thank you for your hint, but the distance between the point and the
>     track is of no relevance for me.
>
>     It seems that the screenshot I posted is a bit misleading
>     (https://i.stack.imgur.com/MqPhK.jpg). Seems I picked a point
>     where the
>     associated trackpoints are also the closest ones. But the
>     information we
>     need is the time, where the observer was when taking the observation.
>     And this might also be on a more distant location.
>
>     Furthermore, I do not intend to create a new layer. I'm just
>     looking for
>     a way to highlight the track point(s) which compare best by time with
>     the selected observation point.
>
>     So, still coulnd't find out how to query points in another layer
>     through
>     an expression/ function, and in case I'll find out, how to compare it
>     with datetime values.
>
>     Cheers,
>
>     Bernd
>
>     Am 16.05.19 um 13:28 schrieb Mike Flannigan:
>     >
>     > Hi,
>     >
>     > I'm thinking this is easier done outside of QGIS, but if you
>     > want to do in within QGIS this is one way:
>     > https://www.qgistutorials.com/en/docs/nearest_neighbor_analysis.html
>     >
>     >
>     > Mike
>     >
>     >
>     > On 5/15/2019 1:17 PM, qgis-user-request at lists.osgeo.org
>     <mailto:qgis-user-request at lists.osgeo.org> wrote:
>     >> Hi folks,
>     >> I'm sure there must be some "easy" solution, but hours of searching
>     >> provided no examples I could learn from to develop a solution:
>     >>
>     >> I have a point layer of bird observations and a GPS-track
>     recorded while
>     >> observing the birds in an area.
>     >> For the interpretation of the findings it would be good to
>     easily find
>     >> out, from which position the recorded observation was taken.
>     >>
>     >> My idea was to compare the timestamp of the bird point with the
>     >> timestamps of the track points, but I run into huge obstacles
>     right at
>     >> start:
>     >> I know that I "could" query across layers by the expression
>     aggregate().
>     >> Unfortunately, the docs on that is so sparse, that I have not the
>     >> slightest idea how to even start with that
>     >>
>     (https://docs.qgis.org/testing/en/docs/user_manual/working_with_vector/expression.html?highlight=expression#aggregates-functions)
>     >>
>     >>
>     >> Ok, in case I could find out how to properly use aggregate, how
>     do I
>     >> best compare those time stamps?
>     >> The layers derive from gpx-files and QGIS identifies the time
>     tag as
>     >> QDateTime.
>     >>
>     >> As it is not very likely that a bird observation was recorded
>     at the
>     >> exact same time as a track point was generated, there is no way to
>     >> actually compare the values, cause the bird time value will
>     always lie
>     >> in between two track point values.
>     >> What kind of query can be used to identify those two trackpoints ?
>     >>
>     >> In short: How can I identify/highligt/mark track points that have a
>     >> timestamp close the selected point in another layer?
>     >>
>     >> I already created a question on this on gis stackexchange with a
>     >> screenshot,but no reactions so far.
>     >>
>     >>
>     >> Thankful for any hint
>     >> Bernd
>     >
>     _______________________________________________
>     Qgis-user mailing list
>     Qgis-user at lists.osgeo.org <mailto:Qgis-user at lists.osgeo.org>
>     List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>     Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-user/attachments/20190518/f8ea15c3/attachment.html>