add quantile regression, change description and documentation #7

HISTORY.rst

@@ -1,11 +1,14 @@
 
+.. _l-HISTORY:
+
 =======
 History
 =======
 
-current - 2018-04-14 - 0.00Mb
+current - 2018-05-07 - 0.00Mb
 =============================
 
+* `7`: add quantile regression (2018-05-07)
 * `5`: replace flake8 by code style (2018-04-14)
 * `1`: change background for cells in notebooks converted into rst then in html, highlight-ipython3 (2018-01-05)
 

_doc/sphinxdoc/source/blog/2017/2017-10-18_first_day.rst

@@ -6,5 +6,5 @@
     :categories: blog
 
     Machine learned models are black boxes.
-    The modules tries to implements some functions
+    The module tries to implements some functions
     to get insights on machine learned models.

_doc/sphinxdoc/source/blog/2018/2018-05-07_quantile_regression.rst

@@ -0,0 +1,18 @@
+
+.. blogpost::
+    :title: Quantile regression with scikit-learn.
+    :keywords: scikit-learn, quantile regression
+    :date: 2018-05-07
+    :categories: machine learning
+
+    :epkg:`scikit-learn` does not have any quantile regression.
+    :epkg:`statsmodels` does have one
+    `QuantReg <http://www.statsmodels.org/dev/generated/statsmodels.regression.quantile_regression.QuantReg.html>`_
+    but I wanted to try something I did for my teachings
+    `Régression Quantile <http://www.xavierdupre.fr/app/ensae_teaching_cs/helpsphinx3/notebooks/td_note_2017_2.html?highlight=mediane>`_
+    based on `Iteratively reweighted least squares <https://en.wikipedia.org/wiki/Iteratively_reweighted_least_squares>`_.
+    I thought it was a good case study to turn a simple algorithm into
+    a learner :epkg:`scikit-learn` can reused in a pipeline.
+    The notebook :ref:`quantileregressionrst` demonstrates it
+    and it is implemented in
+    :class:`QuantileLinearRegression <mlinsights.mlmodel.quantile_regression.QuantileLinearRegression>`.

_doc/sphinxdoc/source/conf.py

@@ -3,7 +3,7 @@
 import os
 import datetime
 import re
-import sphinx_redactor_theme
+import sphinx_rtd_theme
 
 
 sys.path.insert(0, os.path.abspath(os.path.join(os.path.split(__file__)[0])))
@@ -24,7 +24,8 @@
 
 from pyquickhelper.helpgen.default_conf import set_sphinx_variables, get_default_stylesheet
 set_sphinx_variables(__file__, "mlinsights", "Xavier Dupré", 2018,
-                     "sphinx_redactor_theme", sphinx_redactor_theme.get_html_theme_path(),
+                     "sphinx_rtd_theme", [
+                         sphinx_rtd_theme.get_html_theme_path()],
                      locals(), extlinks=dict(
                          issue=('https://github.com/sdpython/mlinsights/issues/%s', 'issue')),
                      title="mlinsights", book=True)
@@ -43,6 +44,16 @@
 
 mathdef_link_only = True
 
+epkg_dictionary['keras'] = 'https://keras.io/'
 epkg_dictionary['Iris'] = 'http://scikit-learn.org/stable/auto_examples/datasets/plot_iris_dataset.html'
 epkg_dictionary['RandomForestRegressor'] = 'http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestRegressor.html'
 epkg_dictionary['REST API'] = "https://en.wikipedia.org/wiki/Representational_state_transfer"
+
+epkg_dictionary.update({
+    'pandas': ('http://pandas.pydata.org/pandas-docs/stable/',
+               ('http://pandas.pydata.org/pandas-docs/stable/generated/pandas.{0}.html', 1),
+               ('http://pandas.pydata.org/pandas-docs/stable/generated/pandas.{0}.{1}.html', 2)),
+    'sklearn': ('http://scikit-learn.org/stable/',
+                ('http://scikit-learn.org/stable/modules/generated/{0}.html', 1),
+                ('http://scikit-learn.org/stable/modules/generated/{0}.{1}.html', 2)),
+})

_unittests/ut_documentation/test_nb_quantile.py

@@ -0,0 +1,65 @@
+# -*- coding: utf-8 -*-
+"""
+@brief      test log(time=13s)
+"""
+
+import sys
+import os
+import unittest
+
+
+try:
+    import pyquickhelper as skip_
+except ImportError:
+    path = os.path.normpath(
+        os.path.abspath(
+            os.path.join(
+                os.path.split(__file__)[0],
+                "..",
+                "..",
+                "..",
+                "pyquickhelper",
+                "src")))
+    if path not in sys.path:
+        sys.path.append(path)
+    import pyquickhelper as skip_
+
+try:
+    import src
+except ImportError:
+    path = os.path.normpath(
+        os.path.abspath(
+            os.path.join(
+                os.path.split(__file__)[0],
+                "..",
+                "..")))
+    if path not in sys.path:
+        sys.path.append(path)
+    import src
+
+from pyquickhelper.loghelper import fLOG
+from pyquickhelper.pycode import add_missing_development_version
+from pyquickhelper.ipythonhelper import test_notebook_execution_coverage
+import src.mlinsights
+
+
+class TestNotebookQuantile(unittest.TestCase):
+
+    def setUp(self):
+        add_missing_development_version(["jyquickhelper"], __file__, hide=True)
+
+    def test_notebook_quantile(self):
+        fLOG(
+            __file__,
+            self._testMethodName,
+            OutputPrint=__name__ == "__main__")
+
+        self.assertTrue(src.mlinsights is not None)
+        folder = os.path.join(os.path.dirname(__file__),
+                              "..", "..", "_doc", "notebooks")
+        test_notebook_execution_coverage(
+            __file__, "quantile", folder, 'mlinsights', fLOG=fLOG)
+
+
+if __name__ == "__main__":
+    unittest.main()

_unittests/ut_mlmodel/test_quantile_regression.py

@@ -0,0 +1,136 @@
+# -*- coding: utf-8 -*-
+"""
+@brief      test log(time=2s)
+"""
+
+import sys
+import os
+import unittest
+import numpy
+import pandas
+from sklearn.linear_model import LinearRegression
+
+
+try:
+    import pyquickhelper as skip_
+except ImportError:
+    path = os.path.normpath(
+        os.path.abspath(
+            os.path.join(
+                os.path.split(__file__)[0],
+                "..",
+                "..",
+                "..",
+                "pyquickhelper",
+                "src")))
+    if path not in sys.path:
+        sys.path.append(path)
+    import pyquickhelper as skip_
+
+
+try:
+    import src
+except ImportError:
+    path = os.path.normpath(
+        os.path.abspath(
+            os.path.join(
+                os.path.split(__file__)[0],
+                "..",
+                "..")))
+    if path not in sys.path:
+        sys.path.append(path)
+    import src
+
+from pyquickhelper.pycode import ExtTestCase, get_temp_folder
+from src.mlinsights.mlmodel import QuantileLinearRegression
+
+
+class TestQuantileRegression(ExtTestCase):
+
+    def test_quantile_regression_no_intercept(self):
+        X = numpy.array([[0.1, 0.2], [0.2, 0.3]])
+        Y = numpy.array([1., 1.1])
+        clr = LinearRegression(fit_intercept=False)
+        clr.fit(X, Y)
+        clq = QuantileLinearRegression(fit_intercept=False)
+        clq.fit(X, Y)
+        self.assertEqual(clr.intercept_, 0)
+        self.assertEqualArray(clr.coef_, clq.coef_)
+        self.assertEqual(clq.intercept_, 0)
+        self.assertEqualArray(clr.intercept_, clq.intercept_)
+
+    def test_quantile_regression_intercept(self):
+        X = numpy.array([[0.1, 0.2], [0.2, 0.3], [0.3, 0.3]])
+        Y = numpy.array([1., 1.1, 1.2])
+        clr = LinearRegression(fit_intercept=True)
+        clr.fit(X, Y)
+        clq = QuantileLinearRegression(verbose=False, fit_intercept=True)
+        clq.fit(X, Y)
+        self.assertNotEqual(clr.intercept_, 0)
+        self.assertNotEqual(clq.intercept_, 0)
+        self.assertEqualArray(clr.intercept_, clq.intercept_)
+        self.assertEqualArray(clr.coef_, clq.coef_)
+
+    def test_quantile_regression_diff(self):
+        X = numpy.array([[0.1], [0.2], [0.3], [0.4]])
+        Y = numpy.array([1., 1.1, 1.2, 10])
+        clr = LinearRegression(fit_intercept=True)
+        clr.fit(X, Y)
+        clq = QuantileLinearRegression(verbose=False, fit_intercept=True)
+        clq.fit(X, Y)
+        self.assertNotEqual(clr.intercept_, 0)
+        self.assertNotEqual(clq.intercept_, 0)
+        self.assertNotEqualArray(clr.coef_, clq.coef_)
+        self.assertNotEqualArray(clr.intercept_, clq.intercept_)
+        self.assertLesser(clq.n_iter_, 10)
+
+    def test_quantile_regression_pandas(self):
+        X = pandas.DataFrame(numpy.array([[0.1, 0.2], [0.2, 0.3]]))
+        Y = numpy.array([1., 1.1])
+        clr = LinearRegression(fit_intercept=False)
+        clr.fit(X, Y)
+        clq = QuantileLinearRegression(fit_intercept=False)
+        clq.fit(X, Y)
+        self.assertEqual(clr.intercept_, 0)
+        self.assertEqualArray(clr.coef_, clq.coef_)
+        self.assertEqual(clq.intercept_, 0)
+        self.assertEqualArray(clr.intercept_, clq.intercept_)
+
+    def test_quantile_regression_list(self):
+        X = [[0.1, 0.2], [0.2, 0.3]]
+        Y = numpy.array([1., 1.1])
+        clq = QuantileLinearRegression(fit_intercept=False)
+        self.assertRaise(lambda: clq.fit(X, Y), TypeError)
+
+    def test_quantile_regression_list(self):
+        X = numpy.random.random(1000)
+        eps1 = (numpy.random.random(900) - 0.5) * 0.1
+        eps2 = numpy.random.random(100) * 2
+        eps = numpy.hstack([eps1, eps2])
+        X = X.reshape((1000, 1))
+        Y = X * 3.4 + 5.6 + eps
+
+        clq = QuantileLinearRegression(verbose=False, fit_intercept=True)
+        self.assertRaise(lambda: clq.fit(X, Y), ValueError)
+
+        Y = X.ravel() * 3.4 + 5.6 + eps
+
+        clq = QuantileLinearRegression(verbose=False, fit_intercept=True)
+        clq.fit(X, Y)
+
+        clr = LinearRegression(fit_intercept=True)
+        clr.fit(X, Y)
+
+        self.assertNotEqual(clr.intercept_, 0)
+        self.assertNotEqual(clq.intercept_, 0)
+        self.assertNotEqualArray(clr.coef_, clq.coef_)
+        self.assertNotEqualArray(clr.intercept_, clq.intercept_)
+        self.assertLesser(clq.n_iter_, 10)
+
+        pr = clr.predict(X)
+        pq = clq.predict(X)
+        self.assertEqual(pr.shape, pq.shape)
+
+
+if __name__ == "__main__":
+    unittest.main()

appveyor.yml

@@ -21,7 +21,7 @@ install:
   - "%PYTHON%\\Scripts\\pymy_install3 --set=pyquickhelper"
   - "%PYTHON%\\Scripts\\pymy_install3 --task=tool --source=zip graphviz"
   - "%PYTHON%\\Scripts\\pip install pyquickhelper --no-deps"
-  - "%PYTHON%\\Scripts\\pip install wcwidth guzzle_sphinx_theme  backcall absl-py"
+  - "%PYTHON%\\Scripts\\pip install wcwidth sphinx_rtd_theme  backcall absl-py"
   - "%PYTHON%\\Scripts\\pymy_install3 scikit-learn scipy tensorflow protobuf h5py"
   - "%PYTHON%\\Scripts\\pip install keras pandas_streaming kiwisolver"
   - "set PATH=%PATH%;C:\\projects\\jyquickhelper\\build\\update_modules\\Graphviz\\bin"

requirements.txt

@@ -4,7 +4,5 @@ guzzle_sphinx_theme
 jyquickhelper
 pandas_streaming
 pyquickhelper>=1.7.2539
-sphinx-redactor-theme
-sphinxjp.themes.basicstrap
-sphinxjp.themes.sphinxjp
+sphinx_rtd_theme
 wheel

setup.py

@@ -18,7 +18,7 @@
 requirements = None
 
 KEYWORDS = project_var_name + ', Xavier Dupré'
-DESCRIPTION = """Look for insights about machine learned models"""
+DESCRIPTION = """Extends scikit-learn with a couple of new models, transform, metrics, plotting."""
 CLASSIFIERS = [
     'Programming Language :: Python :: 3',
     'Intended Audience :: Developers',
@@ -215,7 +215,6 @@ def write_version():
         package_data=package_data,
         # data_files=data_files,
         install_requires=['scikit-learn', 'pandas',
-                          'pillow', 'matplotlib', 'h5py',
-                          'pandas_streaming'],
+                          'matplotlib', 'pandas_streaming'],
         # include_package_data=True,
     )

src/mlinsights/featurizers/ml_featurizer.py

@@ -93,7 +93,7 @@ def wrap_predict_sklearn(X, fct, many):
 
 def model_featurizer_lr(model):
     """
-    Builds a featurizer from a :epkg:`scikit-learn:linear_model:LogisticRegresion`.
+    Builds a featurizer from a :epkg:`scikit-learn:linear_model:LogisticRegression`.
     It returns a function which returns ``model.decision_function(X)``.
 
     @param      model       model to use to featurize a vector

src/mlinsights/mlmodel/__init__.py

@@ -0,0 +1,5 @@
+"""
+@file
+@brief Shortcuts to *mlmodel*.
+"""
+from .quantile_regression import QuantileLinearRegression