Closed mengxr closed 6 years ago
Merging #143 into master will increase coverage by
<.01%
. The diff coverage is100%
.
@@ Coverage Diff @@
## master #143 +/- ##
==========================================
+ Coverage 85.37% 85.37% +<.01%
==========================================
Files 33 34 +1
Lines 1921 1922 +1
Branches 44 41 -3
==========================================
+ Hits 1640 1641 +1
Misses 281 281
Impacted Files | Coverage Δ | |
---|---|---|
...a/com/databricks/sparkdl/DeepImageFeaturizer.scala | 93.65% <100%> (ø) |
:arrow_up: |
...cala/org/apache/spark/sql/types/DataTypeShim.scala | 100% <100%> (ø) |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update a44fcbb...8a832f0. Read the comment docs.
In Spark SQL, nullability is a hint used during optimization and codegen to skip nullchecks, but not intended as an enforcement mechanism or as an implication that null values do exist. It might get dropped through the pipeline.
This PR switches to
DataType.equalsIgnoreNullability
for the check. Without the change, the test would fail with:cc: @jkbradley