search

OutlierDetectionJL / OutlierDetectionPython.jl

Wrapper for Python-based Outlier Detection Algorithms in Julia
MIT License
2 stars 0 forks source link

`whitening` bug in CDDetector #4

Open ablaom opened 10 months ago

ablaom commented 10 months ago 
using Pkg
Pkg.activate(temp=true)
Pkg.add("MLJBase")
Pkg.add("OutlierDetectionPython")

using MLJBase, OutlierDetectionPython

X = (a = rand(10), b = rand(10))
detector = OutlierDetectionPython.CDDetector()

mach = machine(detector, X)
fit!(mach)

# [ Info: Training machine(CDDetector(whitening = true, …), …).
# ┌ Error: Problem fitting the machine machine(CDDetector(whitening = true, …), …).
# └ @ MLJBase ~/.julia/packages/MLJBase/xcS6N/src/machines.jl:683
# [ Info: Running type checks...
# [ Info: Type checks okay.
# ERROR: Python: TypeError: CD.__init__() got an unexpected keyword argument 'whitening'
# Python stacktrace: none
# Stacktrace:
#   [1] pythrow()
#     @ PythonCall ~/.julia/packages/PythonCall/wXfah/src/err.jl:94
(jl_gWxT0t) pkg> st
Status `/private/var/folders/4n/gvbmlhdc8xj973001s6vdyw00000gq/T/jl_gWxT0t/Project.toml`
  [a7f614a8] MLJBase v1.1.0
  [2449c660] OutlierDetectionPython v0.2.2

FYI: You can add generic interface tests using MLJTestInterface.jl, along the lines of this example. These tests would have caught this error.

davnn commented 10 months ago

Looks like I forgot to add this detector to the tests (https://github.com/OutlierDetectionJL/OutlierDetectionPython.jl/blob/master/test/runtests.jl). I'll have a look.

ablaom commented 6 months ago

Any progress here?