Util unit tests #261
Util unit tests #261
Conversation
| from tools.cell_census_builder.util import array_chunker, is_positive_integral, uricat | ||
|
|
||
|
|
||
| def test_is_positive_integral() -> None: |
There was a problem hiding this comment.
To be pedantic, this method should be called is_nonnegative_integral since 0 is an accepted value. If that looks better I can make the change.
There was a problem hiding this comment.
feel free to improve it!
There was a problem hiding this comment.
If you do this, please also add a test of empty array, e.g.,
assert not is_nonnegative_integral(np.array([], dtype=np.float32))
and equivalent for sparse.
Codecov Report
@@ Coverage Diff @@
## main #261 +/- ##
==========================================
+ Coverage 92.22% 93.52% +1.29%
==========================================
Files 33 34 +1
Lines 1943 2023 +80
==========================================
+ Hits 1792 1892 +100
+ Misses 151 131 -20
Flags with carried forward coverage won't be shown. Click here to find out more.
📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more |
|
|
||
|
|
||
| def test_is_positive_integral() -> None: | ||
| X = np.array([1, 2, 3, 4]) |
There was a problem hiding this comment.
these result in an array of dtype np.int64. The function only accepts floats (both float32 and float64), although it does not raise an error if the input violates the type signature.
These tests should explicitly create arrays of dtype float32 or float64. Prefer float32 for most tests, as that is what we typically receive.
| assert chunked[4].nnz == 19 | ||
|
|
||
| # Other formats are rejected by the method | ||
| X = triu(np.ones(100).reshape(10, 10)).tocoo() |
There was a problem hiding this comment.
recommend testing this by passing something other than COO (eg., LIL or BSR), as it is fairly likely we will add COO support the the array_chunker at some point. Just change the final call from tocoo() to tolil() (or anything other than csr/csc/coo)
| X = csr_matrix([[1, 2, 3], [4, 5, 6]]) | ||
| assert is_nonnegative_integral(X) | ||
|
|
||
| X = csr_matrix([[-1, 2, 3], [4, 5, 6]]) |
There was a problem hiding this comment.
these are also int64 - same issue as with ndarray, they need a float dtype to be specified.
There was a problem hiding this comment.
Apologies, for some reason I thought scipy.sparse created float matrices by default. Fixing it.
Unit tests for:
is_positive_integralarray_chunkeruricat