### Apply Trained Model to New Data

Source: https://github.com/arundo/adtk/blob/develop/docs/quickstart.rst

Loads new time series data for testing, validates it, and uses the previously trained SeasonalAD model to detect anomalies in this new data. The results are then visualized.

```python
>>> s_test = pd.read_csv("./testing.csv", index_col="Datetime", parse_dates=True, squeeze=True)
>>> s_test = validate_series(s_test)
>>> print(s_test)
>>> anomalies_pred = seasonal_ad.detect(s_test)
>>> plot(s_test, anomaly=anomalies_pred,
...          ts_linewidth=1, anomaly_color='red', anomaly_tag="marker")
```

--------------------------------

### Install ADTK from Source

Source: https://github.com/arundo/adtk/blob/develop/docs/install.rst

Installs the latest, potentially unstable, version of ADTK from its source code repository. This involves cloning the repository, checking out the develop branch, and then installing locally.

```console
git clone https://github.com/arundo/adtk.git
cd adtk/
git checkout develop
pip install ./
```

--------------------------------

### Cross-check Detection with Known Anomalies

Source: https://github.com/arundo/adtk/blob/develop/docs/quickstart.rst

Loads known anomaly data, converts it into an event format using ADTK's to_events, and visualizes both the known anomalies and the model-detected anomalies on the training data for comparison.

```python
>>> known_anomalies = pd.read_csv("./known_anomalies.csv", index_col="Datetime", parse_dates=True, squeeze=True)
>>> from adtk.data import to_events
>>> known_anomalies = to_events(known_anomalies)
>>> print(known_anomalies)
>>> plot(s_train,
...          anomaly={"Known": known_anomalies, "Model": anomalies},
...          anomaly_tag={"Known": "span", "Model": "marker"},
...          anomaly_color={"Known": "orange", "Model": "red"})
```

--------------------------------

### Install ADTK Documentation Dependencies

Source: https://github.com/arundo/adtk/blob/develop/docs/README.md

Installs the necessary Python packages for building the ADTK documentation from the requirements-docs.txt file.

```bash
$ pip install -r requirements-docs.txt
```

--------------------------------

### Install Documentation Dependencies

Source: https://github.com/arundo/adtk/blob/develop/docs/developer.rst

Installs ADTK with the 'doc' extra, which includes packages necessary for generating the project's documentation using Sphinx.

```console
pip install adtk[doc]
```

--------------------------------

### Install ADTK from Source

Source: https://github.com/arundo/adtk/blob/develop/README.md

Installs the latest, potentially unstable, version of ADTK directly from its source code repository. This method involves cloning the repository, checking out the develop branch, and then installing locally.

```shell
git clone https://github.com/arundo/adtk.git
cd adtk/
git checkout develop
pip install ./
```

--------------------------------

### Install ADTK from PyPI

Source: https://github.com/arundo/adtk/blob/develop/README.md

Installs the latest stable release of the Anomaly Detection Toolkit (ADTK) from the Python Package Index (PyPI). This is the recommended installation method for most users.

```shell
pip install adtk
```

--------------------------------

### Detect Seasonal Pattern Violations

Source: https://github.com/arundo/adtk/blob/develop/docs/quickstart.rst

Initializes and applies the SeasonalAD detector to identify anomalies based on seasonal patterns. It then visualizes the detected anomalies on the time series data.

```python
>>> from adtk.detector import SeasonalAD
>>> seasonal_ad = SeasonalAD()
>>> anomalies = seasonal_ad.fit_detect(s_train)
>>> plot(s_train, anomaly=anomalies, anomaly_color="red", anomaly_tag="marker")
```

--------------------------------

### Visualize Time Series Data

Source: https://github.com/arundo/adtk/blob/develop/docs/quickstart.rst

Visualizes the loaded time series data using ADTK's plot function. This helps in understanding the data's trends and patterns before applying anomaly detection.

```python
>>> from adtk.visualization import plot
>>> plot(s_train)
```

--------------------------------

### Install ADTK from PyPI

Source: https://github.com/arundo/adtk/blob/develop/docs/install.rst

Installs the latest stable release of ADTK from the Python Package Index (PyPI). Requires Python 3.5 or later.

```console
pip install adtk
```

--------------------------------

### Arundo ADTK Project Setup and Tracking

Source: https://github.com/arundo/adtk/blob/develop/docs/releasehistory.rst

Describes package setup options for developers and fixes related to Travis and Coveralls status tracking.

```python
# Added some package setup options for developers
# Fixed an issue of tracking Travis and Coveralls status
```

--------------------------------

### Load and Validate Time Series Data

Source: https://github.com/arundo/adtk/blob/develop/docs/quickstart.rst

Loads time series data from a CSV file using pandas, validates it using ADTK's validate_series function, and prints the resulting series. This step is crucial for ensuring data quality before anomaly detection.

```python
>>> import pandas as pd
>>> s_train = pd.read_csv("./training.csv", index_col="Datetime", parse_dates=True, squeeze=True)
>>> from adtk.data import validate_series
>>> s_train = validate_series(s_train)
>>> print(s_train)
```

--------------------------------

### Install Development Dependencies

Source: https://github.com/arundo/adtk/blob/develop/docs/developer.rst

Installs ADTK with the 'dev' extra, which includes the required formatter (Black) and linter (isort) for development.

```console
pip install adtk[dev]
```

--------------------------------

### Install Testing Dependencies

Source: https://github.com/arundo/adtk/blob/develop/docs/developer.rst

Installs ADTK with the 'test' extra, which includes testing frameworks like pytest, Coverage.py, and tox for automated testing across different Python environments.

```console
pip install adtk[test]
```

--------------------------------

### ADTK Component Overview

Source: https://github.com/arundo/adtk/blob/develop/docs/userguide.rst

This section outlines the core components of ADTK: detectors for scanning time series, transformers for feature extraction, and aggregators for combining detection results. It also introduces Pipelines and Pipenets for composing these components.

```python
import adtk.detector
import adtk.transformer
import adtk.aggregator
import adtk.pipe

# Detectors scan time series for anomalies.
# Transformers transform time series for feature extraction.
# Aggregators combine multiple detection results.

# Pipelines connect components sequentially.
# Pipenets connect components in more complex, non-sequential ways.

# Example of creating a custom component from a function:
# from adtk.detector import CustomizedDetector1D
# def my_detector_function(time_series):
#     # ... anomaly detection logic ...
#     return anomaly_series
# 
# customized_detector = CustomizedDetector1D(my_detector_function)

```

--------------------------------

### Spike and Level Shift Detection with ADTK

Source: https://github.com/arundo/adtk/blob/develop/docs/userguide.rst

Detectors for spikes and level shifts, which are time-dependent anomalies. These often utilize the DoubleRollingAggregate transformer.

```python
from adtk.detector import PersistAD, LevelShiftAD
from adtk.transformer import DoubleRollingAggregate

# Example usage (conceptual):
# double_rolling_transformer = DoubleRollingAggregate(agg_func='mean', window_size=10, agg_func_right='mean', window_size_right=5)
# persist_detector = PersistAD(double_rolling_transformer)
# level_shift_detector = LevelShiftAD(double_rolling_transformer)

```

--------------------------------

### Outlier Detection Detectors in ADTK

Source: https://github.com/arundo/adtk/blob/develop/docs/userguide.rst

Detectors for identifying outliers in time series data. These can use user-defined thresholds or learn normal ranges from historical data.

```python
from adtk.detector import ThresholdAD, QuantileAD, InterQuartileRangeAD, GeneralizedESDTestAD

# Example usage (conceptual):
# threshold_detector = ThresholdAD(lower_threshold=-5, upper_threshold=5)
# quantile_detector = QuantileAD(c=3)
# iqr_detector = InterQuartileRangeAD(c=1.5)
# esd_detector = GeneralizedESDTestAD(alpha=0.05, side='both')

```

--------------------------------

### Rolling Aggregate for Temporal Pattern Detection

Source: https://github.com/arundo/adtk/blob/develop/docs/userguide.rst

Utilize RollingAggregate to detect temporal changes by measuring statistics within sliding windows. Useful for identifying temporary anomalies like high request frequencies.

```python
from adtk.transformer import RollingAggregate

# Example tracking the number of non-zero values in a window
rolling_non_zeros = RollingAggregate(agg=lambda x: (x != 0).sum())

```

--------------------------------

### Univariate and Multivariate Time Series Handling in ADTK

Source: https://github.com/arundo/adtk/blob/develop/docs/userguide.rst

ADTK automatically applies univariate transformers and detectors to each series in a multivariate DataFrame. For anomalies requiring simultaneous analysis of multiple series (e.g., heat index), multivariate transformers and detectors should be considered.

```python
# Applying a univariate transformer to a multivariate DataFrame
# Example: df is a pandas DataFrame with multiple columns
# univariate_transformer.fit_transform(df)

# For multivariate-specific anomalies, use multivariate detectors/transformers
# Example: multivariate_detector.fit_transform(df)

```

--------------------------------

### Seasonal Decomposition with ClassicSeasonalDecomposition

Source: https://github.com/arundo/adtk/blob/develop/docs/userguide.rst

Remove seasonal patterns from time series using ClassicSeasonalDecomposition to highlight deviations from normal seasonal behavior in the residual series. Suitable for data with fixed periods.

```python
from adtk.transformer import ClassicSeasonalDecomposition

# Example decomposing a time series with a weekly period
seasonal_decomposition = ClassicSeasonalDecomposition(period=7)

```

--------------------------------

### Rolling Aggregate Statistics with DoubleRollingAggregate

Source: https://github.com/arundo/adtk/blob/develop/docs/userguide.rst

Detect shifts in time series patterns by applying various statistics like standard deviation over rolling windows. This transformer quantifies patterns of interest for anomaly detection.

```python
from adtk.transformer import DoubleRollingAggregate

# Example using standard deviation to detect volatility shifts
double_rolling_std = DoubleRollingAggregate(agg=lambda x: x.std())

```

--------------------------------

### Autoregression for Cyclic Time Series

Source: https://github.com/arundo/adtk/blob/develop/docs/userguide.rst

Capture changes in autoregressive relationships for cyclic (but not seasonal) time series. Autoregression can be effective when seasonality decomposition is not appropriate due to variable cycle lengths.

```python
from adtk.detector import AutoregressionAD

# Initialize AutoregressionAD detectorautoregression_detector = AutoregressionAD()

```

--------------------------------

### Build ADTK HTML Documentation

Source: https://github.com/arundo/adtk/blob/develop/docs/README.md

Builds the HTML version of the ADTK documentation using the make command. The output will be available in the _build/html directory.

```bash
$ make html
```

--------------------------------

### ADTk Pipenet Parameter and Algorithm Changes (Version 0.5.0)

Source: https://github.com/arundo/adtk/blob/develop/docs/releasehistory.rst

Highlights changes in version 0.5.0, including the modification of the `steps` parameter in pipenet from a list to a dictionary, the addition of a `summary` method to pipenet, and corrections to seasonal decomposition algorithms. It also mentions updates to the demo notebook and plotting functions.

```python
Changed the parameter `steps` of pipenet from list to dict
Added method `summary` to pipenet
Corrected some major algorithmic issues on seasonal decomposition
    - Removed STL decomposition transformer, and hence the corresponding option in SeasonalAD detector
    - Recreated classic seasonal decomposition transformer
Updated the demo notebook in the documentation
Added an option to hide legend in the plotting function
```

--------------------------------

### ADTK Versioning Scheme

Source: https://github.com/arundo/adtk/blob/develop/docs/developer.rst

Details the versioning strategy for ADTK, differentiating between stable releases and pre-release versions. Stable releases follow semantic versioning (MAJOR.MINOR.PATCH), while pre-release versions incorporate pull request information.

```APIDOC
Stable Release Versions:
  - Format: MAJOR.MINOR.PATCH (e.g., 0.Y.Z)
  - Minor version increment (Y): Indicates API-breaking changes (new features, new models).
  - Patch version increment (Z): Indicates non-API-breaking changes (bug fixes, documentation updates).
  - Releases are published to PyPI and GitHub.
  - Stable documentation corresponds to the most recent release.

Pre-release Versions:
  - Generated from pull requests merged into 'master' or 'develop' branches.
  - Format for API-changing PRs: 0.[Y+1].0-dev.N+pr.M
    - N: Monotonic increasing index.
    - M: Pull request index.
    - Example: Merging PR #37 (API change) to 'develop' after 0.1.2 release -> 0.2.0-dev.1+pr.37
  - Format for non-API-changing PRs: 0.Y.[Z+1]-dev.N+pr.M
    - N: Monotonic increasing index.
    - M: Pull request index.
    - Example: Merging PR #38 (bug fix) to 'master' after 0.1.2 release -> 0.1.3-dev.1+pr.38
  - Latest documentation corresponds to the most recent pre-release in the 'develop' branch.
```

--------------------------------

### ADTK Metrics Module

Source: https://github.com/arundo/adtk/blob/develop/docs/api/metrics.rst

This section details the members of the adtk.metrics module, including any inherited members. It serves as a reference for the functionalities available within the metrics component of the ADTK library.

```python
.. automodule:: adtk.metrics
   :members:
   :inherited-members:
```

--------------------------------

### ADTK Transformers Module

Source: https://github.com/arundo/adtk/blob/develop/docs/api/transformers.rst

This section provides an overview of the classes within the `adtk.transformer` module. It includes details on inherited members and their functionalities for time series data transformation.

```python
from adtk.transformer import *

# Example usage of a transformer (e.g., ScalerTransformer)
# from adtk.data import load_dataset
# data = load_dataset("realKnownCause", 
#                     "A1Benchmark", 
#                     "realKnownCause-A1Benchmark.csv")
# 
# scaler = ScalerTransformer()
# scaled_data = scaler.fit_transform(data)
# print(scaled_data.head())
```

--------------------------------

### adtk.pipe Module Members

Source: https://github.com/arundo/adtk/blob/develop/docs/api/pipe.rst

This section documents the members of the adtk.pipe module, including classes and functions for creating anomaly detection pipelines. It details the functionality and usage of these components within the ADTK library.

```python
import adtk.pipe

# Example usage (assuming other ADTK components are imported and configured)
# from adtk.detector import ThresholdAD
# from adtk.data import load_dataset

# data = load_dataset("realKnownCause")
# detector = ThresholdAD(high_value_only=True)
# pipenet = adtk.pipe.Pipenet(data, detector)
# pipenet.fit()
# pipenet.detect()

# pipeline = adtk.pipe.Pipeline()
# pipeline.add(detector)
# pipeline.fit(data)
# pipeline.detect(data)
```

--------------------------------

### Arundo ADTK Version 0.4.1 Changes

Source: https://github.com/arundo/adtk/blob/develop/docs/releasehistory.rst

Details changes in version 0.4.1, including fixes for tox environments and documentation.

```text
Version 0.4.1 (Nov 21, 2019)
===================================
- Fixed an issue of tox environments
- Minor spelling/grammar fix in documentation
```

--------------------------------

### ADTk Unit Tests, Dependency Changes, and Bug Fixes (Version 0.5.1)

Source: https://github.com/arundo/adtk/blob/develop/docs/releasehistory.rst

Details improvements in version 0.5.1, including the addition of new unit tests, removal of seaborn dependency, and fixes for bugs in the metric module, OutlierDetector, transformer pipeline, and pipenet summary.

```python
Added many new unit tests, and modified some old unit test
Removed seaborn from dependencies (use matplotlib built-in style now)
Fixed a bug in the metric module of dict objects as input
Fixed a bug in the detector OutlierDetector that output series has dtype object if NaN is present
Fixed a bug in transformer pipeline that detect and transform methods are confused
Fixed a bug in pipenet that an aggregator node may crash if its input is from a node where subset contains a single item
Fixed a bug in pipenet summary that subset column are always "all" even if not
Some minor optimization of code
```

--------------------------------

### Model Class Inheritance Diagram

Source: https://github.com/arundo/adtk/blob/develop/docs/inheritance.rst

This diagram displays the hierarchical relationships between different model classes. It shows how base classes like _Model, _NonTrainableModel, and _TrainableModel are extended by more specific classes such as univariate/multivariate models, detectors, and transformers, including custom implementations.

```console
_Model
    |-- _NonTrainableModel
            |-- _NonTrainableUnivariateModel
            |       |-- _NonTrainableUnivariateDetector
            |       |       └-- ThresholdAD
            |       |       
            |       └-- _NonTrainableUnivariateTransformer
            |               |-- RollingAggregate
            |               |-- DoubleRollingAggregate
            |               |-- Retrospect
            |               └-- StandardScale
            |       
            |       └-- _NonTrainableMultivariateModel
            |               └-- _NonTrainableMultivariateTransformer
            |                       └-- SumAll
            |
            |-- _TrainableModel
                    |-- _TrainableUnivariateModel
                    |       |-- _TrainableUnivariateDetector
                    |       |       |-- QuantileAD
                    |       |       |-- InterQuartileRangeAD
                    |       |       |-- GeneralizedESDTestAD
                    |       |       |-- PersistAD
                    |       |       |-- LevelShiftAD
                    |       |       |-- VolatilityShiftAD
                    |       |       |-- SeasonalAD
                    |       |       |-- AutoregressionAD
                    |       |       └-- CustomizedDetector1D
                    |       |
                    |       └-- _TrainableUnivariateTransformer
                    |               |-- ClassicSeasonalDecomposition
                    |               └-- CustomizedTransformer1D
                    |
                    └-- _TrainableMultivariateModel
                            |-- _TrainableMultivariateDetector
                            |       |-- MinClusterDetector
                            |       |-- OutlierDetector
                            |       |-- RegressionAD
                            |       |-- PcaAD
                            |       └-- CustomizedDetectorHD
                            |
                            └-- _TrainableMultivariateTransformer
                                    |-- RegressionResidual
                                    |-- PcaProjection
                                    |-- PcaReconstruction
                                    |-- PcaReconstructionError
                                    └-- CustomizedTransformerHD
            |
            └-- _Aggregator
                    |-- AndAggregator
                    |-- OrAggregator
                    └-- CustomizedAggregator
```

--------------------------------

### ADTK Detector Module

Source: https://github.com/arundo/adtk/blob/develop/docs/api/detectors.rst

This section details the detectors available in the `adtk.detector` module. It lists members and inherited members, providing insights into the functionalities and classes related to anomaly detection.

```python
import adtk.detector

# Accessing detector functionalities
# Example: Using a specific detector class (assuming one exists like ThresholdADFDetector)
# from adtk.detector import ThresholdADFDetector
# detector = ThresholdADFDetector(c=2.5, side='both')
# print(detector)
```

--------------------------------

### Arundo ADTK Version 0.3.0 Changes

Source: https://github.com/arundo/adtk/blob/develop/docs/releasehistory.rst

Details the initial release of Arundo ADTK in version 0.3.0.

```text
Version 0.3.0 (Sep 27, 2019)
===================================
- Initial release
```

--------------------------------

### ADTk Module Changes and API Redesign (Version 0.6.0)

Source: https://github.com/arundo/adtk/blob/develop/docs/releasehistory.rst

Details significant changes in version 0.6.0, including API redesigns for visualization, removal of the resample module, enhancements to data expansion functions, changes in output types, and modifications to required model parameters. It also notes the refactoring of inheritance structures and the addition of Python 3.8 support.

```python
Re-designed the API of :py:mod:`adtk.visualization.plot`
Removed :py:mod:`adtk.data.resample` because its functionality is highly overlapped with pandas resampler module
Made :py:mod:`adtk.data.expand_events` accept events in the form of pandas Series/DataFrame
Made :py:mod:`adtk.data.expand_events` accept time delta in the form of `str` or `int`
Changed the output type of :py:mod:`adtk.data.split_train_test` from a 2-tuple of lists to a list of 2-tuples
Turned the following model parameters required from optional
    - `window` in :py:mod:`adtk.detector.LevelShiftAD`
    - `window` in :py:mod:`adtk.detector.VolatilityShiftAD`
    - `window` in :py:mod:`adtk.transformer.RollingAggregate`
    - `window` in :py:mod:`adtk.transformer.DoubleRollingAggregate`
    - `model` in :py:mod:`adtk.detector.MinClusterDetector`
    - `model` in :py:mod:`adtk.detector.OutlierDetector`
    - `target` and `regressor` in :py:mod:`adtk.detector.RegressionAD`
    - `target` and `regressor` in :py:mod:`adtk.transformer.RegressionResidual`
    - `aggregate_func` in :py:mod:`adtk.aggregator.CustomizedAggregator`
    - `detect_func` in :py:mod:`adtk.detector.CustomizedDetector1D`
    - `detect_func` in :py:mod:`adtk.detector.CustomizedDetectorHD`
    - `transform_func` in :py:mod:`adtk.transformer.CustomizedTransformer1D`
    - `transform_func` in :py:mod:`adtk.detector.CustomizedTransformer1D`
    - `steps` in :py:mod:`adtk.pipe.Pipeline`
Added consistency check between training and testing inputs in multivariate models
Improved time index check in time-dependent models
Turned all second-order sub-modules private, and a user now can only import from the following first-order modules
    - :py:mod:`adtk.detector`
    - :py:mod:`adtk.transformer`
    - :py:mod:`adtk.aggregator`
    - :py:mod:`adtk.pipe`
    - :py:mod:`adtk.data`
    - :py:mod:`adtk.metrics`
    - :py:mod:`adtk.visualization`
Refactored the inheritance structure of model components (see :ref:`inheritance`)
Added Python 3.8 support
Fixed compatibility issues with statsmodels v0.11
Fixed compatibility issues with pandas v1.0
Created an interactive demo notebook in Binder
Added type hints, and added type checking in CI/CD test
Added `Black` and `isort` to developer requirement and CI/CD check
Optimized release process by publishing package to PyPI through GitHub Actions
Improved docstrings and API documentation
Fixed many minor bugs and typos
```

--------------------------------

### ADTk Workflow Optimization and Bug Fixes (Version 0.5.4)

Source: https://github.com/arundo/adtk/blob/develop/docs/releasehistory.rst

Focuses on optimizing the workflow for applying univariate models to pandas DataFrames in version 0.5.4. This includes adding more informative error messages, fixing bugs related to model-column matching, and clarifying the documentation.

```python
Optimized the workflow of how a univariate model is applied to pandas DataFrame
    - Added more informative error messages
    - Fixed some bugs resulting in model-column matching error due to inconsistency between output Series names and DataFrame columns
    - Clarified the workflow in the documentation
```

--------------------------------

### Arundo ADTK Version 0.4.0 Changes

Source: https://github.com/arundo/adtk/blob/develop/docs/releasehistory.rst

Details changes in version 0.4.0, including Python 3.5 support, improved tests, and README updates.

```text
Version 0.4.0 (Nov 18, 2019)
===================================
- Added support to Python 3.5
- Better unit tests on dependencies
- Minor typo fix in documentation
- Minor code optimization
- Added download statistics to README
- Added coverage test
```

--------------------------------

### ADTk Dependency Management and Bug Fixes (Version 0.5.3)

Source: https://github.com/arundo/adtk/blob/develop/docs/releasehistory.rst

Addresses compatibility issues with statsmodels in version 0.5.3 by implementing a hotfix to avoid errors caused by statsmodels v0.11. This involves requiring statsmodels dependency to be less than version 0.11.

```python
Quick hotfix to avoid errors caused by statsmodels v0.11 by requiring statsmodels dependency <0.11
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.