Describe the bug
pyscamp.selfjoin returns a zero vector if m=3. If m>= 4 the results are correct.
To Reproduce
x = np.random.rand(10)
pyscamp.selfjoin(x, m=3)
gives a zero vector like this:
(array([0., 0., 0., 0., 0., 0., 0., 0.], dtype=float32),
array([0, 1, 2, 3, 4, 5, 6, 7], dtype=int32))
stumpy gives correct results:
mparray([[0.543935636387785, 4, -1, 4],
[0.6665659559814567, 5, -1, 5],
[0.29289592419909743, 7, 0, 7],
[1.5383853098049423, 1, 1, 6],
[0.543935636387785, 0, 0, 7],
[0.6665659559814567, 1, 1, 7],
[0.6135462405979734, 2, 2, -1],
[0.29289592419909743, 2, 2, -1]], dtype=object)
Outputs of pyscamp.selfjoin and stumpy.stump are identical if m>=4.
Expected behavior
Output of pyscamp.selfjoin should match stumpy.stump.
System Information
- OS: Rocky Linux 8.10
- System Architecture: x86_64
- Package and Version: pyscamp 4.0.1 installed from source using pip
Describe the bug
pyscamp.selfjoin returns a zero vector if m=3. If m>= 4 the results are correct.
To Reproduce
gives a zero vector like this:
stumpy gives correct results:
Outputs of pyscamp.selfjoin and stumpy.stump are identical if m>=4.
Expected behavior
Output of pyscamp.selfjoin should match stumpy.stump.
System Information