Opened 8 years ago

Closed 8 years ago

#1205 closed defect (fixed)

WCPS1.5_Wrong geo-referencing from WCPS query to RASQL query

Reported by: Bang Pham Huu Owned by:
Priority: major Milestone: 9.3
Component: petascope Version: development
Keywords: WCPS 2.0, wrong georeferencing. Cc: Dimitar Misev, Vlad Merticariu, Alex Dumitru
Complexity: Medium

Description

test wcps query "82-timestamp_in_subset.test"

for c in (eobstest) return encode(
  (c[
    t("1950-01-02"),
    Long(20:35),
    Lat(10:15)]),
"tiff", "nodata=0")

which is translated wrong geo-referencing in WCPS 2.0 (in here it get all the bounding box instead of trimming).

SELECT encode( ( c[1,-10:19,121:130] ),
"GTiff" ,
"nodata=0;xmin=25.0;xmax=75.5;ymin=-40.5;ymax=75.5;crs=EPSG:4326") FROM eobstest AS c

instead of correct translation as WCPS 1.0

select encode(c[1,-10:20,121:131],
"GTiff",
"nodata=0;xmin=20.0;xmax=35.0;ymin=10.0;ymax=15.0;crs=EPSG:4326") from eobstest AS c where oid(c)=46081

Change History (3)

comment:1 by Bang Pham Huu, 8 years ago

Summary: WCPS2.0_Wrong geo-referencing from WCPS query to RASQL queryWCPS1.5_Wrong geo-referencing from WCPS query to RASQL query

comment:2 by Bang Pham Huu, 8 years ago

In WCPS 1.0 there is a bug when using "*" in axis, with WCPS query like this

for c in (eobstest) return encode(c[t(0), Lat(0:*), Long(0:35)], "tiff")

will returns Rasql

select encode(c[0,-50:20,0:151], "GTiff") from eobstest AS c where oid(c)=1537

without GeoReferencing parameters (xmin, xmax, ymin, ymax)

While with same WCPS query with same limit upper bound but using number

for c in (eobstest) return encode(c[t(0), Lat(0:75.5), Long(0:35)], "tiff")

it will returns the GeoReferencing parameters as

select encode(c[0,-50:20,0:151], "GTiff",
 "xmin=0.0;xmax=35.0;ymin=0.0;ymax=75.5;crs=EPSG:4326") from eobstest AS c where oid(c)=1537

This problem will be fixed in WCPS1.5 with this ticket.

comment:3 by Bang Pham Huu, 8 years ago

Milestone: 10.09.3
Resolution: fixed
Status: newclosed

It is now set correct bounding box according to the domain of coverage's axis, thanks for Vlad's idea, close ticket.

Note: See TracTickets for help on using tickets.