Layer 1: Tabular Anomaly Detection
====================================

PyOD has 43 tabular detectors covering probabilistic, linear, proximity, ensemble, and deep learning approaches. All use the same ``fit``/``predict``/``decision_function`` API.

.. code-block:: python

    from pyod.models.iforest import IForest
    clf = IForest()
    clf.fit(X_train)
    y_train_scores = clf.decision_scores_
    y_test_scores = clf.decision_function(X_test)

----

Recommended Starting Points
----------------------------

Based on `ADBench <https://github.com/Minqi824/ADBench>`__ (NeurIPS 2022, 57 datasets, 30 algorithms):

* `ECOD <https://github.com/yzhao062/pyod/blob/master/examples/ecod_example.py>`__ -- parameter-free, highly interpretable, top ADBench performance
* `IForest <https://github.com/yzhao062/pyod/blob/master/examples/iforest_example.py>`__ -- tree ensemble, scales to high dimensions
* `KNN <https://github.com/yzhao062/pyod/blob/master/examples/knn_example.py>`__ -- proximity-based, good baseline
* `LOF <https://github.com/yzhao062/pyod/blob/master/examples/lof_example.py>`__ -- density-based, good for local anomalies
* `COPOD <https://github.com/yzhao062/pyod/blob/master/examples/copod_example.py>`__ -- copula-based, fast

----

All Tabular Examples
--------------------

**Probabilistic:** `ECOD <https://github.com/yzhao062/pyod/blob/master/examples/ecod_example.py>`__, `COPOD <https://github.com/yzhao062/pyod/blob/master/examples/copod_example.py>`__, `ABOD <https://github.com/yzhao062/pyod/blob/master/examples/abod_example.py>`__, `MAD <https://github.com/yzhao062/pyod/blob/master/examples/mad_example.py>`__, `SOS <https://github.com/yzhao062/pyod/blob/master/examples/sos_example.py>`__, `QMCD <https://github.com/yzhao062/pyod/blob/master/examples/qmcd_example.py>`__, `KDE <https://github.com/yzhao062/pyod/blob/master/examples/kde_example.py>`__, `Sampling <https://github.com/yzhao062/pyod/blob/master/examples/sampling_example.py>`__, `GMM <https://github.com/yzhao062/pyod/blob/master/examples/gmm_example.py>`__

**Linear Models:** `PCA <https://github.com/yzhao062/pyod/blob/master/examples/pca_example.py>`__, `KPCA <https://github.com/yzhao062/pyod/blob/master/examples/kpca_example.py>`__, `MCD <https://github.com/yzhao062/pyod/blob/master/examples/mcd_example.py>`__, `CD <https://github.com/yzhao062/pyod/blob/master/examples/cd_example.py>`__, `OCSVM <https://github.com/yzhao062/pyod/blob/master/examples/ocsvm_example.py>`__, `LMDD <https://github.com/yzhao062/pyod/blob/master/examples/lmdd_example.py>`__

**Proximity-Based:** `LOF <https://github.com/yzhao062/pyod/blob/master/examples/lof_example.py>`__, `COF <https://github.com/yzhao062/pyod/blob/master/examples/cof_example.py>`__, `CBLOF <https://github.com/yzhao062/pyod/blob/master/examples/cblof_example.py>`__, `LOCI <https://github.com/yzhao062/pyod/blob/master/examples/loci_example.py>`__, `HBOS <https://github.com/yzhao062/pyod/blob/master/examples/hbos_example.py>`__, `HDBSCAN <https://github.com/yzhao062/pyod/blob/master/examples/hdbscan_example.py>`__, `KNN <https://github.com/yzhao062/pyod/blob/master/examples/knn_example.py>`__, `SOD <https://github.com/yzhao062/pyod/blob/master/examples/sod_example.py>`__, `ROD <https://github.com/yzhao062/pyod/blob/master/examples/rod_example.py>`__

**Outlier Ensembles:** `IForest <https://github.com/yzhao062/pyod/blob/master/examples/iforest_example.py>`__, `INNE <https://github.com/yzhao062/pyod/blob/master/examples/inne_example.py>`__, `DIF <https://github.com/yzhao062/pyod/blob/master/examples/dif_example.py>`__, `Feature Bagging <https://github.com/yzhao062/pyod/blob/master/examples/feature_bagging_example.py>`__, `LSCP <https://github.com/yzhao062/pyod/blob/master/examples/lscp_example.py>`__, `XGBOD <https://github.com/yzhao062/pyod/blob/master/examples/xgbod_example.py>`__, `LODA <https://github.com/yzhao062/pyod/blob/master/examples/loda_example.py>`__, `SUOD <https://github.com/yzhao062/pyod/blob/master/examples/suod_example.py>`__

**Neural Networks:** `AutoEncoder <https://github.com/yzhao062/pyod/blob/master/examples/auto_encoder_example.py>`__, `VAE <https://github.com/yzhao062/pyod/blob/master/examples/vae_example.py>`__, `DeepSVDD <https://github.com/yzhao062/pyod/blob/master/examples/deepsvdd_example.py>`__, `SO_GAAL <https://github.com/yzhao062/pyod/blob/master/examples/so_gaal_example.py>`__, `MO_GAAL <https://github.com/yzhao062/pyod/blob/master/examples/mo_gaal_example.py>`__, AnoGAN, `ALAD <https://github.com/yzhao062/pyod/blob/master/examples/alad_example.py>`__, `AE1SVM <https://github.com/yzhao062/pyod/blob/master/examples/ae1svm_example.py>`__, `DevNet <https://github.com/yzhao062/pyod/blob/master/examples/devnet_example.py>`__

----

Example Walkthrough
-------------------

Full example: `knn_example.py <https://github.com/yzhao062/pyod/blob/master/examples/knn_example.py>`__

1. Import and generate data:

.. code-block:: python

    from pyod.models.knn import KNN
    from pyod.utils.data import generate_data, evaluate_print

    contamination = 0.1
    X_train, X_test, y_train, y_test = generate_data(
        n_train=200, n_test=100, contamination=contamination)

2. Fit and predict:

.. code-block:: python

    clf = KNN()
    clf.fit(X_train)

    y_train_pred = clf.labels_                  # 0: inlier, 1: outlier
    y_train_scores = clf.decision_scores_       # raw scores
    y_test_pred = clf.predict(X_test)
    y_test_scores = clf.decision_function(X_test)

3. Evaluate:

.. code-block:: python

    evaluate_print('KNN', y_test, y_test_scores)
    # KNN ROC:0.9989, precision @ rank n:0.9