5094
87%
develop: 89%

Ran 18 Dec 2018 01:05AM UTC

Jobs 1

Files 99

Run time 12s

Badge

Embed ▾

pending completion

Build # 5094

Build Type

push

travis-ci

Committed by

Kirill888

Commit Message

Make sure src/dst roi are padded properly in all situtations

One of the old equivalence tests were failing due to one row/col of destination
image were not being populated by the new code. Even though those pixels do not
overlap with the source image they are being filled by rasterio.reproject when
called on the whole src/dst image.

That equivalence tests is still failing when using cubic interpolation, but it
no longer fails when using average, so it's a step up. I suspect that cubic has
some sensitivity to the location of x,y coordinates, so if you were to translate
both source and destination regions by the same amount you would get slightly
different numeric results for relative x,y coordinates and from that slightly
different pixel values. I have reviewed result in the notebook and it's really
just a numerical accuracy issue.

I had to change hash values in the ingestion test again and to pad out requested
GeoBox by one more pixel to get nodata values. This test needs re-think, but
also I need to look into the data being produced and see if it makes sense in a
first place. So far it just tests that results are repeatable across
runs/machines for a given reprojection approach.

Run Details

7288 of 8373 relevant lines covered (87.04%)

0.87 hits per line

Jobs

ID	Job ID	Ran	Files	Coverage
2	5094.2	18 Dec 2018 01:06AM UTC	0	87.04	Travis Job 5094.2

Source Files on build 5094

Coverage	∆	File	Lines	Relevant	Covered	Missed	Hits/Line

opendatacube / datacube-core / 5094
87%
develop: 89%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 5094

opendatacube / datacube-core / 5094 87% develop: 89%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 5094

opendatacube / datacube-core / 5094
87%
develop: 89%

README BADGES
x