Basic structure

Author: jhosoume

Makefile

@@ -0,0 +1,30 @@
+PACKAGE := NoiseFiltersPy
+
+all: clean install-complete code-check
+.PHONE: all clean code-check pypi install-complete install
+
+clean:
+	find . -name "*.so" -o -name "*.pyc" -o -name "*.md5" -o -name "*.pyd" -o -name "*~" | xargs rm -f
+	rm -rf dist
+	rm -rf build
+	rm -rf .pytest_cache/
+	rm -rf .mypy_cache/
+
+code-check:
+	flake8 $(PACKAGE)
+	pylint $(PACKAGE) -j 0 -d 'C0103, R0913, R0902, R0914, C0302, R0904, R0801, E1101, C0330, E1136'
+	mypy $(PACKAGE) --ignore-missing-imports
+
+pypi: clean ## Send the package to pypi.
+	pip install -U twine wheel
+	python3 setup.py sdist bdist_wheel
+	twine upload dist/*
+
+install-complete: 
+	pip install -U -e .
+	pip install -U -r requirements.txt
+	pip install -U -r requirements-dev.txt
+	pip install -U -r requirements-docs.txt
+
+install:
+	pip install .

NoiseFiltersPy/CNN.py

@@ -6,7 +6,7 @@
 
 
 class CNN:
-    def __init__(self, max_neighbours = 5):
+    def __init__(self, max_neighbours: int = 5):
         self.max_neighbours = max_neighbours
         self.filter = Filter(parameters = {})
         self.clf = KNeighborsClassifier(n_neighbors = 1, n_jobs = -1)

NoiseFiltersPy/DROP.py

@@ -1,6 +1,7 @@
 from sklearn.model_selection import KFold
 from sklearn.neighbors import KNeighborsClassifier
 import numpy as np
+import typing as t
 
 from NoiseFiltersPy.Filter import *
 

NoiseFiltersPy/ENN.py

@@ -1,15 +1,16 @@
 from sklearn.model_selection import KFold
 from sklearn.neighbors import KNeighborsClassifier
 import numpy as np
+import typing as t
 
 from NoiseFiltersPy.Filter import *
 
 class ENN:
-    def __init__(self, neighbours = 3):
+    def __init__(self, neighbours: int = 3):
         self.neighbours = neighbours
         self.filter = Filter(parameters = {"neighbours": self.neighbours})
 
-    def __call__(self, data, classes):
+    def __call__(self, data: t.Sequence, classes: t.Sequence) -> Filter:
         self.isNoise = np.array([False] * len(classes))
         self.clf = KNeighborsClassifier(n_neighbors = self.neighbours, algorithm = 'kd_tree', n_jobs = -1)
         for indx in range(len(data)):

NoiseFiltersPy/Filter.py

@@ -1,9 +1,12 @@
+import typing as t
+
 class Filter():
-    def __init__(self, parameters):
+    def __init__(self, parameters: t.Dict):
         # Removed Indexes
-        self.remIndx = []
-        self.repIndx = []
+        self.remIndx: t.List = []
         self.parameters = parameters
-    def set_cleanData(self, data, classes):
+
+    def set_cleanData(self, data: t.Sequence, classes: t.Sequence) -> t.NoReturn:
+        # Helper function to set data without noise
         self.cleanData = data
         self.cleanClasses = classes

NoiseFiltersPy/HARF.py

@@ -1,13 +1,14 @@
 from sklearn.model_selection import KFold
 from sklearn.ensemble import RandomForestClassifier
 import numpy as np
+import typing as t
 
 from NoiseFiltersPy.Filter import *
 
 
 class HARF:
-    def __init__(self, nfolds = 10, agreementLevel = 0.7,
-                 ntrees = 500, seed = 0):
+    def __init__(self, nfolds: int = 10, agreementLevel: float = 0.7,
+                 ntrees: int = 500, seed: int = 0):
         # Some data verification
         # Data can be a DataFrame or a Numpy Array
         if (agreementLevel < 0.5 or agreementLevel > 1):
@@ -22,7 +23,7 @@ def __init__(self, nfolds = 10, agreementLevel = 0.7,
         self.clf = RandomForestClassifier(n_estimators = ntrees, random_state = seed, n_jobs = -1)
         self.filter = Filter(parameters = {"nfolds": self.nfolds, "ntrees": self.ntrees, "agreementLevel": self.agreementLevel})
 
-    def __call__(self, data, classes):
+    def __call__(self, data: t.Sequence, classes: t.Sequence) -> Filter:
         self.splits = self.k_fold.split(data)
         self.isNoise = np.array([False] * len(classes))
         for train_indx, test_indx in self.splits:

NoiseFiltersPy/TomekLinks.py

@@ -1,13 +1,13 @@
 import numpy as np
+import typing as t
 
 from NoiseFiltersPy.Filter import *
 
-
 class TomekLinks:
     def __init__(self):
         self.filter = Filter(parameters = {})
 
-    def __call__(self, data, classes):
+    def __call__(self, data: t.Sequence, classes: t.Sequence) -> Filter:
         levels = list(set(classes))
         classes = np.array(classes)
         class1Indxes = np.argwhere(classes == levels[0])

examples/aenn_example.py

@@ -6,3 +6,4 @@
 classes = dataset.target
 aenn = AENN()
 filter = aenn(data, classes)
+print(filter.cleanData)

examples/cnn_example.py

@@ -1,4 +1,6 @@
 from sklearn import datasets
+
+import context
 from NoiseFiltersPy.CNN import CNN
 
 dataset = datasets.load_iris()

examples/context.py

@@ -0,0 +1,3 @@
+import os
+import sys
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))

examples/18_mfeat-morphological.arff examples/datasets/18_mfeat-morphological.arff

@@ -0,0 +1,6 @@
+pytest
+pytest-cov
+pytest-xdist
+mypy
+flake8
+pylint

requirements-dev.txt

@@ -0,0 +1,6 @@
+sphinx
+sphinx_rtd_theme
+pillow
+seaborn
+sphinx-gallery
+numpydoc

requirements-docs.txt

@@ -0,0 +1,3 @@
+numpy
+scipy
+scikit-learn

requirements.txt

@@ -3,9 +3,31 @@
 with open("README.md", "r") as fh:
     long_description = fh.read()
 
+INSTALL_REQUIRES = []
+
+EXTRAS_REQUIRE = {
+    'code-check': [
+        'pytest',
+        'mypy',
+        'flake8',
+        'pylint'
+    ],
+    'tests': [
+        'pytest',
+        'pytest-cov',
+        'pytest-xdist',
+    ],
+    'docs': [
+        'sphinx',
+        'sphinx-gallery',
+        'sphinx_rtd_theme',
+        'numpydoc',
+    ]
+}
+
 setuptools.setup(
     name="NoiseFiltersPy",
-    version="0.0.5",
+    version="0.0.8",
     author="Juliana Hosoume and Luis Faina",
     author_email="[email protected]",
     description="Python implementation of NoiseFiltersR",
@@ -18,5 +40,11 @@
         "License :: OSI Approved :: MIT License",
         "Operating System :: OS Independent",
     ],
+    install_requires=[
+        'scikit-learn>=0.24',
+        'numpy',
+        'scipy',
+        'pandas'
+    ],
     python_requires='>=3.5',
 )