Opened 11 years ago
Closed 11 years ago
#475 closed defect (fixed)
Coverage iterator over double/date intervals
Reported by: | Dimitar Misev | Owned by: | Dimitar Misev |
---|---|---|---|
Priority: | major | Milestone: | 9.0 |
Component: | petascope | Version: | development |
Keywords: | Cc: | Piero Campalani, Dirk Daems, Peter Baumann | |
Complexity: | Medium |
Description
It would be very useful if the coverage iterator in a coverage constructor can iterate over double and date intervals, so that we can iterate over geo and time coordinates. This should be possible to implement fairly easily. E.g.
for c in (Cov) return coverage newcov over px x(15.5:93.432) pt t(2013-09-10:2013-09-25) values ...
Change History (3)
comment:1 by , 11 years ago
Cc: | added |
---|---|
Owner: | changed from | to
Status: | new → accepted |
comment:2 by , 11 years ago
I submitted patch to allow imageCrsDomain
. An example test case in the systemtest is:
for c in (eobstest) return encode( coverage newcov over $px x(imageCrsDomain(c[Lat(25:35)], Lat)), $pt t(imageCrsDomain(c, t)) values avg(c[Lat($px),t($pt)]) , "csv")
comment:3 by , 11 years ago
Resolution: | → fixed |
---|---|
Status: | accepted → closed |
Note:
See TracTickets
for help on using tickets.
Ok I take back the "fairly easy" remark.
It would be fairly easy if the over clause was bound to an existing coverage, otherwise it would be necessary to provide a resolution parameter as well for the iterator, and a CRS.
So something like this (0.5 is resolution)
but then there is the issue of translating that to rasql and relating it to coverages in the values clause..
All in all, quite complicated. I think for most purposes it would suffice to allow imageCrsDomain in the over, which would provide the binding to an existing coverage, i.e. something like
much like we made it possible to use
imageCrsDomain
in scale (cf. #387)Thoughts?