32 enhance sphinx bibtext

README.rst

@@ -49,31 +49,32 @@ Detector List
 
 Menelaus implements the following drift detectors.
 
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Type              | Detector                                                       | Abbreviation  | Streaming  | Batch  | Ref.    |
-+===================+================================================================+===============+============+========+=========+
-| Change detection  | Cumulative Sum Test                                            | CUSUM         | x          |        | [C1]_   |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Change detection  | Page-Hinkley                                                   | PH            | x          |        | [C2]_   |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Concept drift     | ADaptive WINdowing                                             | ADWIN         | x          |        | [C3]_   |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Concept drift     | Drift Detection Method                                         | DDM           | x          |        | [C4]_   |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Concept drift     | Early Drift Detection Method                                   | EDDM          | x          |        | [C5]_   |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Concept drift     | Linear Four Rates                                              | LFR           | x          |        | [C6]_   |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Concept drift     | Statistical Test of Equal Proportions to Detect concept drift  | STEPD         | x          |        | [C7]_   |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Data drift        | Confidence Distribution Batch Detection                        | CDBD          |            | x      | [C8]_   |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Data drift        | Hellinger Distance Drift Detection Method                      | HDDDM         |            | x      | [C9]_   |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Data drift        | kdq-Tree Detection Method                                      | kdq-Tree      | x          | x      | [C10]_  |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
-| Data drift        | PCA-Based Change Detection                                     | PCA-CD        | x          |        | [C11]_  |
-+-------------------+----------------------------------------------------------------+---------------+------------+--------+---------+
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Type              | Detector                                                       | Abbreviation  | Streaming  | Batch  | Ref.                             |
++===================+================================================================+===============+============+========+==================================+
+| Change detection  | Cumulative Sum Test                                            | CUSUM         | x          |        | :cite:t:`hinkley1971inference`   |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Change detection  | Page-Hinkley                                                   | PH            | x          |        | :cite:t:`page1954continuous`     |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Concept drift     | ADaptive WINdowing                                             | ADWIN         | x          |        | :cite:t:`bifet2007learning`      |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Concept drift     | Drift Detection Method                                         | DDM           | x          |        | :cite:t:`gama2004learning`       |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Concept drift     | Early Drift Detection Method                                   | EDDM          | x          |        | :cite:t:`baena2006early`         |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Concept drift     | Linear Four Rates                                              | LFR           | x          |        | :cite:t:`wang2015concept`        |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Concept drift     | Statistical Test of Equal Proportions to Detect concept drift  | STEPD         | x          |        | :cite:t:`nishida2007detecting`   |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Data drift        | Confidence Distribution Batch Detection                        | CDBD          |            | x      | :cite:t:`lindstrom2013drift`     |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Data drift        | Hellinger Distance Drift Detection Method                      | HDDDM         |            | x      | :cite:t:`ditzler2011hellinger`   |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Data drift        | kdq-Tree Detection Method                                      | kdq-Tree      | x          | x      | :cite:t:`dasu2006information`    |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+| Data drift        | PCA-Based Change Detection                                     | PCA-CD        | x          |        | :cite:t:`qahtan2015pca`          |
++-------------------+----------------------------------------------------------------+---------------+------------+--------+----------------------------------+
+
 
 
 The three main types of detector are described below. More details can be found 

docs/source/conf.py

@@ -33,6 +33,7 @@
     "sphinx.ext.autodoc",
     "sphinx.ext.napoleon",
     "sphinx.ext.viewcode",
+    "sphinxcontrib.bibtex",
 ]
 
 autodoc_default_options = {
@@ -42,6 +43,9 @@
     "special-members": "__init__",
 }
 
+bibtex_bibfiles = ["refs.bib"]
+bibtex_reference_style = "author_year"
+
 # Add any paths that contain templates here, relative to this directory.
 templates_path = ["_templates"]
 

docs/source/references.rst

@@ -2,14 +2,4 @@
 
 References
 ==================================
-.. [C1] Hinkley, David V. "Inference about the change-point from cumulative sum tests." Biometrika 58.3 (1971): 509-523.
-.. [C2] Page, Ewan S. "Continuous inspection schemes." Biometrika 41.1/2 (1954): 100-115.
-.. [C3] Bifet, Albert, and Ricard Gavalda. "Learning from time-changing data with adaptive windowing." Proceedings of the 2007 SIAM international conference on data mining. Society for Industrial and Applied Mathematics, 2007.
-.. [C4] Gama, Joao, et al. "Learning with drift detection." Brazilian symposium on artificial intelligence. Springer, Berlin, Heidelberg, 2004.
-.. [C5] Baena-Garcıa, Manuel, et al. "Early drift detection method." Fourth international workshop on knowledge discovery from data streams. Vol. 6. 2006.
-.. [C6] Wang, Heng, and Zubin Abraham. "Concept drift detection for streaming data." 2015 international joint conference on neural networks (IJCNN). IEEE, 2015.
-.. [C7] Nishida, Kyosuke, and Koichiro Yamauchi. "Detecting concept drift using statistical testing." International conference on discovery science. Springer, Berlin, Heidelberg, 2007.
-.. [C8] Lindstrom, Patrick, Brian Mac Namee, and Sarah Jane Delany. "Drift detection using uncertainty distribution divergence." Evolving Systems 4.1 (2013): 13-25.
-.. [C9] Ditzler, Gregory, and Robi Polikar. "Hellinger distance based drift detection for nonstationary environments." 2011 IEEE symposium on computational intelligence in dynamic and uncertain environments (CIDUE). IEEE, 2011.
-.. [C10] Dasu, Tamraparni, et al. "An information-theoretic approach to detecting changes in multi-dimensional data streams." In Proc. Symp. on the Interface of Statistics, Computing Science, and Applications. 2006.
-.. [C11] Qahtan, Abdulhakim A., et al. "A pca-based change detection framework for multidimensional data streams: Change detection in multidimensional data streams." Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2015.
+.. bibliography::

docs/source/refs.bib

@@ -0,0 +1,102 @@
+@article{hinkley1971inference,
+  title={Inference about the change-point from cumulative sum tests},
+  author={Hinkley, David V},
+  journal={Biometrika},
+  volume={58},
+  number={3},
+  pages={509--523},
+  year={1971},
+  publisher={Oxford University Press}
+}
+
+@article{page1954continuous,
+  title={Continuous inspection schemes},
+  author={Page, Ewan S},
+  journal={Biometrika},
+  volume={41},
+  number={1/2},
+  pages={100--115},
+  year={1954},
+  publisher={JSTOR}
+}
+
+@inproceedings{bifet2007learning,
+  title={Learning from time-changing data with adaptive windowing},
+  author={Bifet, Albert and Gavalda, Ricard},
+  booktitle={Proceedings of the 2007 SIAM international conference on data mining},
+  pages={443--448},
+  year={2007},
+  organization={SIAM}
+}
+
+@inproceedings{gama2004learning,
+  title={Learning with drift detection},
+  author={Gama, Joao and Medas, Pedro and Castillo, Gladys and Rodrigues, Pedro},
+  booktitle={Brazilian symposium on artificial intelligence},
+  pages={286--295},
+  year={2004},
+  organization={Springer}
+}
+
+@inproceedings{baena2006early,
+  title={Early drift detection method},
+  author={Baena-Garc{\i}a, Manuel and del Campo-{\'A}vila, Jos{\'e} and Fidalgo, Ra{\'u}l and Bifet, Albert and Gavalda, R and Morales-Bueno, Rafael},
+  booktitle={Fourth international workshop on knowledge discovery from data streams},
+  volume={6},
+  pages={77--86},
+  year={2006}
+}
+
+@inproceedings{wang2015concept,
+  title={Concept drift detection for streaming data},
+  author={Wang, Heng and Abraham, Zubin},
+  booktitle={2015 international joint conference on neural networks (IJCNN)},
+  pages={1--9},
+  year={2015},
+  organization={IEEE}
+}
+
+@inproceedings{nishida2007detecting,
+  title={Detecting concept drift using statistical testing},
+  author={Nishida, Kyosuke and Yamauchi, Koichiro},
+  booktitle={International conference on discovery science},
+  pages={264--269},
+  year={2007},
+  organization={Springer}
+}
+
+@article{lindstrom2013drift,
+  title={Drift detection using uncertainty distribution divergence},
+  author={Lindstrom, Patrick and Mac Namee, Brian and Delany, Sarah Jane},
+  journal={Evolving Systems},
+  volume={4},
+  number={1},
+  pages={13--25},
+  year={2013},
+  publisher={Springer}
+}
+
+@inproceedings{ditzler2011hellinger,
+  title={Hellinger distance based drift detection for nonstationary environments},
+  author={Ditzler, Gregory and Polikar, Robi},
+  booktitle={2011 IEEE symposium on computational intelligence in dynamic and uncertain environments (CIDUE)},
+  pages={41--48},
+  year={2011},
+  organization={IEEE}
+}
+
+@inproceedings{dasu2006information,
+  title={An information-theoretic approach to detecting changes in multi-dimensional data streams},
+  author={Dasu, Tamraparni and Krishnan, Shankar and Venkatasubramanian, Suresh and Yi, Ke},
+  booktitle={In Proc. Symp. on the Interface of Statistics, Computing Science, and Applications},
+  year={2006},
+  organization={Citeseer}
+}
+
+@inproceedings{qahtan2015pca,
+  title={A pca-based change detection framework for multidimensional data streams: Change detection in multidimensional data streams},
+  author={Qahtan, Abdulhakim A and Alharbi, Basma and Wang, Suojin and Zhang, Xiangliang},
+  booktitle={Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining},
+  pages={935--944},
+  year={2015}
+}

setup.cfg

@@ -51,6 +51,7 @@ dev =
    sphinx
    sphinx-rtd-theme
    sphinx-autoapi
+   sphinxcontrib-bibtex
 
 [options.packages.find]
 where=src

src/menelaus/change_detection/cusum.py

@@ -9,7 +9,7 @@ class CUSUM(DriftDetector):
     single model performance metric, or could be applied to the mean of a
     feature variable of interest.
 
-    Ref. [C1]_
+    Ref. :cite:t:`hinkley1971inference`
 
     Attributes:
         total_updates (int): number of samples the drift detector has ever

src/menelaus/change_detection/page_hinkley.py

@@ -21,7 +21,7 @@ class PageHinkley(DriftDetector):
     If the threshold is too small, PH may result in many false alarms. If too
     large, the PH test will be more robust, but may miss true drift.
 
-    Ref. [C2]_
+    Ref. :cite:t:`page1954continuous`
 
     Attributes:
         total_updates (int): number of samples the drift detector has ever

src/menelaus/concept_drift/adwin.py

@@ -34,7 +34,7 @@ class ADWIN(DriftDetector):
     When drift occurs, the index of the element at the beginning of ADWIN's new
     window is stored in ``self.retraining_recs``.
 
-    Ref. [C3]_
+    Ref. :cite:t:`bifet2007learning`
 
     Attributes:
         total_updates (int): number of samples the drift detector has ever

src/menelaus/concept_drift/ddm.py

@@ -20,7 +20,7 @@ class DDM(DriftDetector):
     The index of the first sample which triggered a warning/drift state
     (relative to ``self.updates_since_reset``) is stored in ``self.retraining_recs``.
 
-    Ref. [C4]_
+    Ref. :cite:t:`gama2004learning`
 
     Attributes:
         total_updates (int): number of samples the drift detector has ever

src/menelaus/concept_drift/eddm.py

@@ -24,7 +24,7 @@ class EDDM(DriftDetector):
     The index of the first sample which triggered a warning/drift state
     (relative to ``self.updates_since_reset``) is stored in ``self.retraining_recs``.
 
-    Ref. [C5]_
+    Ref. :cite:t:`baena2006early`
 
     Attributes:
         total_updates (int): number of samples the drift detector has ever

src/menelaus/concept_drift/lfr.py

@@ -23,7 +23,7 @@ class LinearFourRates(DriftDetector):
     of number of time steps and estimated empirical rate, if a given combination
     has been simulated before, the bounds are re-used.
 
-    Ref. [C6]_
+    Ref. :cite:t:`wang2015concept`
 
     Attributes:
         total_updates (int): number of samples the drift detector has ever

src/menelaus/concept_drift/stepd.py

@@ -31,7 +31,7 @@ class STEPD(DriftDetector):
     using only a single data point vs. being required to retrain on the entire
     set.
 
-    Ref. [C7]_
+    Ref. :cite:t:`nishida2007detecting`
 
     Attributes:
         total_updates (int): number of samples the drift detector has ever

src/menelaus/data_drift/cdbd.py

@@ -45,7 +45,7 @@ class CDBD(HistogramDensityMethod):
            drift is not detected. The reference batch is updated to include this
            most recent test batch. All statistics are maintained.
 
-    Ref. [C8]_
+    Ref. :cite:t:`lindstrom2013drift`
 
     Attributes:
             total_updates (int): number of batches the drift detector has ever

src/menelaus/data_drift/hdddm.py

@@ -105,7 +105,7 @@ class HDDDM(HistogramDensityMethod):
               calculated using the first test batch, allowing for detection of
               drift on this batch.
 
-    Ref. [C9]_
+    Ref. :cite:t:`ditzler2011hellinger`
 
     Attributes:
         total_updates (int): number of batches the drift detector has ever

src/menelaus/data_drift/histogram_density_method.py

@@ -119,7 +119,7 @@ class HistogramDensityMethod(DriftDetector):
               and samples since reset will be number of batches passed to HDM
               plus 1, due to splitting of reference batch
 
-    Ref. [C8]_ and [C9]_
+    Ref. :cite:t:`lindstrom2013drift` and :cite:t:`ditzler2011hellinger`
 
     Attributes:
         total_updates (int): number of batches the drift detector has ever

src/menelaus/data_drift/kdq_tree.py

@@ -51,7 +51,7 @@ class KdqTree(DriftDetector):
     Note that the current implementation does not explicitly handle categorical
     data.
 
-    Ref. [C10]_
+    Ref. :cite:t:`dasu2006information`
 
 
     Attributes:
@@ -299,18 +299,18 @@ def to_plotly_dataframe(self, tree_id1="build", tree_id2="test", max_depth=None)
 
                 * ``name``: a label corresponding to which feature this split is on
                 * ``idx``: a unique ID for the node, to pass
-                ``plotly.express.treemap``'s id argument
+                    ``plotly.express.treemap``'s id argument
                 * ``parent_idx``: the ID of the node's parent
                 * ``cell_count``: how many samples are in this node in the
-                  reference tree.
+                    reference tree.
                 * ``depth``: how deep the node is in the tree
                 * ``count_diff``: if ``tree_id2`` is specified, the change in
-                  counts from the reference tree.
+                    counts from the reference tree.
                 * ``kss``: the Kulldorff Spatial Scan Statistic for this node,
-                  defined as the Kullback-Leibler divergence for this node
-                  between the reference and test trees, using the individual
-                  node and all other nodes combined as the bins for the
-                  distributions.
+                    defined as the Kullback-Leibler divergence for this node
+                    between the reference and test trees, using the individual
+                    node and all other nodes combined as the bins for the
+                    distributions.
         """
 
         return self._kdqtree.to_plotly_dataframe(tree_id1, tree_id2, max_depth)

src/menelaus/data_drift/pca_cd.py

@@ -24,7 +24,7 @@ class PCACD(DriftDetector):
     Once drift is detected, the reference window is replaced with the current
     test window, and the test window is initialized.
 
-    Ref. [C11]_
+    Ref. :cite:t:`qahtan2015pca`
 
     Attributes:
         total_updates (int): number of samples the drift detector has ever
@@ -56,7 +56,7 @@ def __init__(
         """
         Args:
             window_size (int): size of the reference window. Note that
-                ``PCA_CD``will only try to detect drift periodically, either
+                ``PCA_CD`` will only try to detect drift periodically, either
                 every 100 observations or 5% of the ``window_size``, whichever
                 is smaller.
             ev_threshold (float, optional): Threshold for percent explained