Skip to main content

Python SDK for Optimizely Feature Experimentation, Optimizely Full Stack (legacy), and Optimizely Rollouts.

Project description

Optimizely Feature Experimentation is A/B testing and feature management for product development teams. Experiment in any application. Make every feature on your roadmap an opportunity to learn. Learn more at https://www.optimizely.com/products/experiment/feature-experimentation/ or see our documentation at https://docs.developers.optimizely.com/experimentation/v4.0.0-full-stack/docs/welcome. # Optimizely Python SDK

PyPI version Build Status Coverage Status Apache 2.0

This repository houses the Python SDK for use with Optimizely Feature Experimentation and Optimizely Full Stack (legacy).

Optimizely Feature Experimentation is an A/B testing and feature management tool for product development teams that enables you to experiment at every step. Using Optimizely Feature Experimentation allows for every feature on your roadmap to be an opportunity to discover hidden insights. Learn more at Optimizely.com, or see the developer documentation.

Optimizely Rollouts is free feature flags for development teams. You can easily roll out and roll back features in any application without code deploys, mitigating risk for every feature on your roadmap.

Get Started

Refer to the Python SDK's developer documentation for detailed instructions on getting started with using the SDK.

Requirements

Version 4.0+: Python 3.7+, PyPy 3.7+

Version 3.0+: Python 2.7+, PyPy 3.4+

Install the SDK

The SDK is available through PyPi.

To install:

pip install optimizely-sdk

Feature Management Access

To access the Feature Management configuration in the Optimizely dashboard, please contact your Optimizely customer success manager.

Use the Python SDK

Initialization

You can initialize the Optimizely instance in three ways: with a datafile, by providing an sdk_key, or by providing an implementation of BaseConfigManager. Each method is described below.

  1. Initialize Optimizely with a datafile. This datafile will be used as the source of ProjectConfig throughout the life of Optimizely instance:

    optimizely.Optimizely(
      datafile
    )
    
  2. Initialize Optimizely by providing an 'sdk_key'. This will initialize a PollingConfigManager that makes an HTTP GET request to the URL (formed using your provided sdk key and the default datafile CDN URL template) to asynchronously download the project datafile at regular intervals and update ProjectConfig when a new datafile is received. A hard-coded datafile can also be provided along with the sdk_key that will be used initially before any update:

    optimizely.Optimizely(
      sdk_key='put_your_sdk_key_here'
    )
    

    If providing a datafile, the initialization will look like:

    optimizely.Optimizely(
      datafile=datafile,
      sdk_key='put_your_sdk_key_here'
    )
    
  3. Initialize Optimizely by providing a ConfigManager that implements BaseConfigManager. You may use our PollingConfigManager or AuthDatafilePollingConfigManager as needed:

    optimizely.Optimizely(
      config_manager=custom_config_manager
    )
    

PollingConfigManager

The PollingConfigManager asynchronously polls for datafiles from a specified URL at regular intervals by making HTTP requests.

polling_config_manager = PollingConfigManager(
    sdk_key=None,
    datafile=None,
    update_interval=None,
    url=None,
    url_template=None,
    logger=None,
    error_handler=None,
    notification_center=None,
    skip_json_validation=False
)

Note: You must provide either the sdk_key or URL. If you provide both, the URL takes precedence.

sdk_key The sdk_key is used to compose the outbound HTTP request to the default datafile location on the Optimizely CDN.

datafile You can provide an initial datafile to bootstrap the ProjectConfigManager so that it can be used immediately. The initial datafile also serves as a fallback datafile if HTTP connection cannot be established. The initial datafile will be discarded after the first successful datafile poll.

update_interval The update_interval is used to specify a fixed delay in seconds between consecutive HTTP requests for the datafile.

url The target URL from which to request the datafile.

url_template A string with placeholder {sdk_key} can be provided so that this template along with the provided sdk key is used to form the target URL.

You may also provide your own logger, error_handler, or notification_center.

AuthDatafilePollingConfigManager

The AuthDatafilePollingConfigManager implements PollingConfigManager and asynchronously polls for authenticated datafiles from a specified URL at regular intervals by making HTTP requests.

auth_datafile_polling_config_manager = AuthDatafilePollingConfigManager(
    datafile_access_token,
    *args,
    **kwargs
)

Note: To use AuthDatafilePollingConfigManager, you must create a secure environment for your project and generate an access token for your datafile.

datafile_access_token The datafile_access_token is attached to the outbound HTTP request header to authorize the request and fetch the datafile.

Advanced configuration

The following properties can be set to override the default configurations for PollingConfigManager and AuthDatafilePollingConfigManager.

Property Name Default Value Description
sdk_key None Optimizely project SDK key
datafile None Initial datafile, typically sourced from a local cached source
update_interval 5 minutes Fixed delay between fetches for the datafile
url None Custom URL location from which to fetch the datafile
url_template PollingConfigManager:
https://cdn.optimizely.com/datafiles/{sdk_key}.json
AuthDatafilePollingConfigManager:
https://config.optimizely.com/datafiles/auth/{sdk_key}.json
Parameterized datafile URL by SDK key

A notification signal will be triggered whenever a new datafile is fetched and Project Config is updated. To subscribe to these notifications, use:

notification_center.add_notification_listener(NotificationTypes.OPTIMIZELY_CONFIG_UPDATE, update_callback)

For Further details see the Optimizely Feature Experimentation documentation to learn how to set up your first Python project and use the SDK.

SDK Development

Building the SDK

Build and install the SDK with pip, using the following command:

pip install -e .

Unit Tests

Running all tests

To get test dependencies installed, use a modified version of the install command:

pip install -e '.[test]'

You can run all unit tests with:

pytest

Running all tests in a file

To run all tests under a particular test file you can use the following command:

pytest tests.<file_name_without_extension>

For example, to run all tests under test_event_builder, the command would be:

pytest tests/test_event_builder.py

Running all tests under a class

To run all tests under a particular class of tests you can use the following command:

pytest tests/<file_name_with_extension>::ClassName

For example, to run all tests under test_event_builder.EventTest, the command would be:

pytest tests/test_event_builder.py::EventTest

Running a single test

To run a single test you can use the following command:

pytest tests/<file_name_with_extension>::ClassName::test_name

For example, to run test_event_builder.EventTest.test_init, the command would be:

pytest tests/test_event_builder.py::EventTest::test_init

Contributing

Please see CONTRIBUTING.

Credits

This software incorporates code from the following open source projects:

requests (Apache-2.0 License: https://github.com/psf/requests/blob/master/LICENSE)

pyOpenSSL (Apache-2.0 License https://github.com/pyca/pyopenssl/blob/main/LICENSE)

cryptography (Apache-2.0 https://github.com/pyca/cryptography/blob/main/LICENSE.APACHE)

idna (BSD 3-Clause License https://github.com/kjd/idna/blob/master/LICENSE.md)

Other Optimizely SDKs

Optimizely Python SDK Changelog

5.0.0-beta

Apr 28th, 2023

New Features

The 5.0.0-beta release introduces a new primary feature, Advanced Audience Targeting enabled through integration with Optimizely Data Platform (ODP) (#395, #398, #402, #403, #405).

You can use ODP, a high-performance Customer Data Platform (CDP), to easily create complex real-time segments (RTS) using first-party and 50+ third-party data sources out of the box. You can create custom schemas that support the user attributes important for your business, and stitch together user behavior done on different devices to better understand and target your customers for personalized user experiences. ODP can be used as a single source of truth for these segments in any Optimizely or 3rd party tool.

With ODP accounts integrated into Optimizely projects, you can build audiences using segments pre-defined in ODP. The SDK will fetch the segments for given users and make decisions using the segments. For access to ODP audience targeting in your Feature Experimentation account, please contact your Optimizely Customer Success Manager.

This version includes the following changes:

  • New API added to OptimizelyUserContext:

    • fetchQualifiedSegments(): this API will retrieve user segments from the ODP server. The fetched segments will be used for audience evaluation. The fetched data will be stored in the local cache to avoid repeated network delays.
    • When an OptimizelyUserContext is created, the SDK will automatically send an identify request to the ODP server to facilitate observing user activities.
  • New APIs added to OptimizelyClient:

    • sendOdpEvent(): customers can build/send arbitrary ODP events that will bind user identifiers and data to user profiles in ODP.

For details, refer to our documentation pages:

Breaking Changes

  • ODPManager in the SDK is enabled by default. Unless an ODP account is integrated into the Optimizely projects, most ODPManager functions will be ignored. If needed, ODPManager can be disabled when OptimizelyClient is instantiated.
  • BaseConfigManager abstract class now requires a get_sdk_key method. (#413)
  • PollingConfigManager requires either the sdk_key parameter or datafile containing an sdkKey. (#413)
  • Asynchronous BatchEventProcessor is now the default event processor. (#378)

4.1.1

March 10th, 2023

We updated our README.md and other non-functional code to reflect that this SDK supports both Optimizely Feature Experimentation and Optimizely Full Stack. (#420)

4.1.0

July 7th, 2022

Bug Fixes

  • Fix invalid datafile returned from ProjectConfig.to_datafile and OptimizelyConfig.get_datafile (#321, #384)

4.0.0

January 12th, 2022

New Features

  • Add a set of new APIs for overriding and managing user-level flag, experiment and delivery rule decisions. These methods can be used for QA and automated testing purposes. They are an extension of the OptimizelyUserContext interface (#361, #365, #369):

    • setForcedDecision
    • getForcedDecision
    • removeForcedDecision
    • removeAllForcedDecisions
  • For details, refer to our documentation pages: OptimizelyUserContext and Forced Decision methods.

Breaking Changes:

  • Support for Python v3.4 has been dropped as of this release due to a security vulnerability with PyYAML <v5.4. (#366)
  • We no longer support Python v2.7, v3.5, and v3.6 including PyPy as of this release. (#377)
  • We now support Python v3.7 and above including PyPy3.

3.10.0

September 16th, 2021

New Features

Bug Fixes

3.9.1

July 14th, 2021

Bug Fixes:

3.9.0

June 1st, 2021

New Features

  • Added support for multiple concurrent prioritized experiments per flag. #322

3.8.0

February 12th, 2021

New Features

  • New Features Introducing a new primary interface for retrieving feature flag status, configuration and associated experiment decisions for users (#309). The new OptimizelyUserContext class is instantiated with create_user_context and exposes the following APIs to get OptimizelyDecision:

    • set_attribute
    • decide
    • decide_all
    • decide_for_keys
    • track_event

For details, refer to our documentation page: https://docs.developers.optimizely.com/full-stack/v4.0/docs/python-sdk.

3.7.1

November 19th, 2020

Bug Fixes:

  • Added "enabled" field to decision metadata structure. #306

3.7.0

November 2nd, 2020

New Features

  • Added support for upcoming application-controlled introduction of tracking for non-experiment Flag decisions. #300

3.6.0

October 1st, 2020

New Features:

  • Version targeting using semantic version syntax. #293
  • Datafile accessor API added to access current config as a JSON string. #283

Bug Fixes:

  • Fixed package installation for Python 3.4 and pypy. #298

3.5.2

July 14th, 2020

Bug Fixes:

  • Fixed handling of network and no status code errors when polling for datafile in PollingConfigManager and AuthDatafilePollingConfigManager. (#287)

3.5.1

July 10th, 2020

Bug Fixes:

  • Fixed HTTP request exception handling in PollingConfigManager. (#285)

3.5.0

July 9th, 2020

New Features:

  • Introduced 2 APIs to interact with feature variables:
    • get_feature_variable_json allows you to get value for JSON variables related to a feature.
    • get_all_feature_variables gets values for all variables under a feature.
  • Added support for fetching authenticated datafiles. AuthDatafilePollingConfigManager is a new config manager that allows you to poll for a datafile belonging to a secure environment. You can create a client by setting the datafile_access_token.

Bug Fixes:

  • Fixed log messages for targeted rollouts evaluation. (#268)

3.4.2

June 11th, 2020

Bug Fixes:

  • Adjusted log level for audience evaluation logs. (#267)

3.4.1

March 19th, 2020

Bug Fixes:

3.4.0

January 27th, 2020

New Features:

  • Added a new API to get project configuration static data.
    • Call get_optimizely_config() to get a snapshot of project configuration static data.
    • It returns an OptimizelyConfig instance which includes a datafile revision number, all experiments, and feature flags mapped by their key values.
    • Added caching for get_optimizely_config() - OptimizelyConfig object will be cached and reused for the lifetime of the datafile.
    • For details, refer to our documentation page: https://docs.developers.optimizely.com/full-stack/docs/optimizelyconfig-python.

3.3.1

December 16th, 2019

Bug Fixes:

  • Fixed installation issue on Windows. (#224)
  • Fixed batch event processor deadline reset issue. (#227)
  • Added more batch event processor debug messages. (#227)

3.3.0

October 28th, 2019

New Features:

  • Added support for event batching via the event processor.
    • Events generated by methods like activate, track, and is_feature_enabled will be held in a queue until the configured batch size is reached, or the configured flush interval has elapsed. Then, they will be batched into a single payload and sent to the event dispatcher.
    • To configure event batching, set the batch_size and flush_interval properties when initializing instance of BatchEventProcessor.
    • Event batching is disabled by default. You can pass in instance of BatchEventProcessor when creating Optimizely instance to enable event batching.
    • Users can subscribe to LogEvent notification to be notified of whenever a payload consisting of a batch of user events is handed off to the event dispatcher to send to Optimizely's backend.
  • Introduced blocking timeout in PollingConfigManager. By default, calls to get_config will block for maximum of 10 seconds until config is available.

Bug Fixes:

  • Fixed incorrect log message when numeric metric is not used. (#217)

3.2.0

August 27th, 2019

New Features:

  • Added support for automatic datafile management via PollingConfigManager:
    • The PollingConfigManager is an implementation of the BaseConfigManager.
    • Users may provide one of datafile or SDK key (sdk_key) or both to optimizely.Optimizely. Based on that, the SDK will use the StaticConfigManager or the PollingConfigManager. Refer to the README for more instructions.
    • An initial datafile can be provided to the PollingConfigManager to bootstrap before making HTTP requests for the hosted datafile.
    • Requests for the datafile are made in a separate thread and are scheduled with fixed delay.
    • Configuration updates can be subscribed to by adding the OPTIMIZELY_CONFIG_UPDATE notification listener.
  • Introduced Optimizely.get_feature_variable API. (#191)

Deprecated:

  • NotificationCenter.clear_notifications is deprecated as of this release. Please use NotificationCenter.clear_notification_listeners. (#182)
  • NotificationCenter.clear_all_notifications is deprecated as of this release. Please use NotificationCenter.clear_all_notification_listeners. (#182)

3.2.0b1

July 26th, 2019

New Features:

  • Added support for automatic datafile management via PollingConfigManager:
    • The PollingConfigManager is an implementation of the BaseConfigManager.
    • Users may provide one of datafile or SDK key (sdk_key) or both to optimizely.Optimizely. Based on that, the SDK will use the StaticConfigManager or the PollingConfigManager. Refer to the README for more instructions.
    • An initial datafile can be provided to the PollingConfigManager to bootstrap before making HTTP requests for the hosted datafile.
    • Requests for the datafile are made in a separate thread and are scheduled with fixed delay.
    • Configuration updates can be subscribed to by adding the OPTIMIZELY_CONFIG_UPDATE notification listener.
  • Introduced Optimizely.get_feature_variable API. (#191)

Deprecated:

  • NotificationCenter.clear_notifications is deprecated as of this release. Please use NotificationCenter.clear_notification_listeners. (#182)
  • NotificationCenter.clear_all_notifications is deprecated as of this release. Please use NotificationCenter.clear_all_notification_listeners. (#182)

3.1.0

May 3rd, 2019

New Features:

  • Introduced Decision notification listener to be able to record:
    • Variation assignments for users activated in an experiment.
    • Feature access for users.
    • Feature variable value for users.

Bug Fixes:

  • Feature variable APIs now return default variable value when featureEnabled property is false. (#171)

Deprecated:

  • Activate notification listener is deprecated as of this release. Recommendation is to use the new Decision notification listener. Activate notification listener will be removed in the next major release.

3.0.0

March 1st, 2019

The 3.0 release improves event tracking and supports additional audience targeting functionality.

New Features:

  • Event tracking:
    • The track method now dispatches its conversion event unconditionally, without first determining whether the user is targeted by a known experiment that uses the event. This may increase outbound network traffic.
    • In Optimizely results, conversion events sent by 3.0 SDKs don't explicitly name the experiments and variations that are currently targeted to the user. Instead, conversions are automatically attributed to variations that the user has previously seen, as long as those variations were served via 3.0 SDKs or by other clients capable of automatic attribution, and as long as our backend actually received the impression events for those variations.
    • Altogether, this allows you to track conversion events and attribute them to variations even when you don't know all of a user's attribute values, and even if the user's attribute values or the experiment's configuration have changed such that the user is no longer affected by the experiment. As a result, you may observe an increase in the conversion rate for previously-instrumented events. If that is undesirable, you can reset the results of previously-running experiments after upgrading to the 3.0 SDK. - This will also allow you to attribute events to variations from other Optimizely projects in your account, even though those experiments don't appear in the same datafile.
    • Note that for results segmentation in Optimizely results, the user attribute values from one event are automatically applied to all other events in the same session, as long as the events in question were actually received by our backend. This behavior was already in place and is not affected by the 3.0 release.
  • Support for all types of attribute values, not just strings.
    • All values are passed through to notification listeners.
    • Strings, booleans, and valid numbers are passed to the event dispatcher and can be used for Optimizely results segmentation. A valid number is a finite float or numbers.Integral in the inclusive range [-2 ^ 53, 2 ^ 53].
    • Strings, booleans, and valid numbers are relevant for audience conditions.
  • Support for additional matchers in audience conditions:
    • An exists matcher that passes if the user has a non-null value for the targeted user attribute and fails otherwise.
    • A substring matcher that resolves if the user has a string value for the targeted attribute.
      • gt (greater than) and lt (less than) matchers that resolve if the user has a valid number value for the targeted attribute. A valid number is a finite float or numbers.Integral in the inclusive range [-2 ^ 53, 2 ^ 53].
      • The original (exact) matcher can now be used to target booleans and valid numbers, not just strings.
  • Support for A/B tests, feature tests, and feature rollouts whose audiences are combined using "and" and "not" operators, not just the "or" operator.
  • Datafile-version compatibility check: The SDK will remain uninitialized (i.e., will gracefully fail to activate experiments and features) if given a datafile version greater than 4.
  • Updated Pull Request template and commit message guidelines.

Breaking Changes:

  • Conversion events sent by 3.0 SDKs don't explicitly name the experiments and variations that are currently targeted to the user, so these events are unattributed in raw events data export. You must use the new results export to determine the variations to which events have been attributed.
  • Previously, notification listeners were only given string-valued user attributes because only strings could be passed into various method calls. That is no longer the case. You may pass non-string attribute values, and if you do, you must update your notification listeners to be able to receive whatever values you pass in.

Bug Fixes:

  • Experiments and features can no longer activate when a negatively targeted attribute has a missing, null, or malformed value.
    • Audience conditions (except for the new exists matcher) no longer resolve to false when they fail to find an legitimate value for the targeted user attribute. The result remains null (unknown). Therefore, an audience that negates such a condition (using the "not" operator) can no longer resolve to true unless there is an unrelated branch in the condition tree that itself resolves to true.
  • Updated the default event dispatcher to log an error if the request resolves to HTTP 4xx or 5xx. (#140)
  • All methods now validate that user IDs are strings and that experiment keys, feature keys, feature variable keys, and event keys are non-empty strings.

2.1.1

August 21st, 2018

  • Fix: record conversions for all experiments using an event when using track(#136).

2.1.0

July 2nd, 2018

  • Introduced support for bot filtering (#121).
  • Overhauled logging to use standard Python logging (#123).

2.0.1

June 19th, 2018

  • Fix: send impression event for Feature Test when Feature is disabled (#128).

2.0.0

April 12th, 2018

This major release introduces APIs for Feature Management. It also introduces some breaking changes listed below.

New Features

  • Introduced the is_feature_enabled API to determine whether to show a feature to a user or not.
    is_enabled = optimizel_client.is_feature_enabled('my_feature_key', 'my_user', user_attributes)
  • All enabled features for the user can be retrieved by calling:
    enabled_features = optimizely_client.get_enabled_features('my_user', user_attributes)
  • Introduced Feature Variables to configure or parameterize a feature. There are four variable types: String, Integer, Double, Boolean.
    string_variable = optimizely_client.get_feature_variable_string('my_feature_key', 'string_variable_key', 'my_user')
    integer_variable = optimizely_client.get_feature_variable_integer('my_feature_key', 'integer_variable_key', 'my_user')
    double_variable = optimizely_client.get_feature_variable_double('my_feature_key', 'double_variable_key', 'my_user')
    boolean_variable = optimizely_client.get_feature_variable_boolean('my_feature_key', 'boolean_variable_key', 'my_user')

Breaking changes

  • The track API with revenue value as a stand-alone parameter has been removed. The revenue value should be passed in as an entry in the event tags dict. The key for the revenue tag is revenue and the passed in value will be treated by Optimizely as the value for computing results.
    event_tags = {
      'revenue': 1200
    }

    optimizely_client.track('event_key', 'my_user', user_attributes, event_tags)

2.0.0b1

March 29th, 2018

This beta release introduces APIs for Feature Management. It also introduces some breaking changes listed below.

New Features

  • Introduced the is_feature_enabled API to determine whether to show a feature to a user or not.
    is_enabled = optimizel_client.is_feature_enabled('my_feature_key', 'my_user', user_attributes)
  • All enabled features for the user can be retrieved by calling:
    enabled_features = optimizely_client.get_enabled_features('my_user', user_attributes)
  • Introduced Feature Variables to configure or parameterize a feature. There are four variable types: String, Integer, Double, Boolean.
    string_variable = optimizely_client.get_feature_variable_string('my_feature_key', 'string_variable_key', 'my_user')
    integer_variable = optimizely_client.get_feature_variable_integer('my_feature_key', 'integer_variable_key', 'my_user')
    double_variable = optimizely_client.get_feature_variable_double('my_feature_key', 'double_variable_key', 'my_user')
    boolean_variable = optimizely_client.get_feature_variable_boolean('my_feature_key', 'boolean_variable_key', 'my_user')

Breaking changes

  • The track API with revenue value as a stand-alone parameter has been removed. The revenue value should be passed in as an entry in the event tags dict. The key for the revenue tag is revenue and the passed in value will be treated by Optimizely as the value for computing results.
    event_tags = {
      'revenue': 1200
    }

    optimizely_client.track('event_key', 'my_user', user_attributes, event_tags)

1.4.0

  • Added support for IP anonymization.
  • Added support for notification listeners.
  • Added support for bucketing ID.
  • Updated mmh3 to handle installation failures on Windows 10.

1.3.0

  • Introduced support for forced bucketing.
  • Introduced support for numeric metrics.
  • Updated event builder to support new endpoint.

1.2.1

  • Removed older feature flag parsing.

1.2.0

  • Added user profile service.

1.1.1

  • Updated datafile parsing to be able to handle additional fields.
  • Deprecated Classic project support.

1.1.0

  • Included datafile revision information in log events.
  • Added event tags to track API to allow users to pass in event metadata.
  • Deprecated the event_value parameter from the track method. Should use event_tags to pass in event value instead.
  • Updated event logging endpoint to logx.optimizely.com.

1.0.0

  • Introduced support for Full Stack projects in Optimizely X. No breaking changes from previous version.
  • Introduced more graceful exception handling in instantiation and core methods.
  • Updated whitelisting to precede audience matching.

0.1.3

  • Added support for v2 endpoint and datafile.
  • Updated dispatch_event to consume an Event object instead of url and params. The Event object comprises of four properties: url (string representing URL to dispatch event to), params (dict representing the params to be set for the event), http_verb (one of 'GET' or 'POST') and headers (header values to be sent along).
  • Fixed issue with tracking events for experiments in groups.

0.1.2

  • Updated requirements file.

0.1.1

  • Introduced option to skip JSON schema validation.

0.1.0

  • Beta release of the Python SDK for server-side testing.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

optimizely-sdk-5.0.0b0.tar.gz (119.4 kB view hashes)

Uploaded Source

Built Distribution

optimizely_sdk-5.0.0b0-py3-none-any.whl (154.5 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page