Real-Time Audiences for Feature Experimentation segment qualification methods for the Python SDK
Prerequisites
You must enable and configure Real-Time Audiences for Feature Experimentation before fetching qualified segments and checking if the user is qualified for the given audience segment.
fetch_qualified_segments
Minimum SDK version
5.0.0
Description
You can use the fetch_qualified_segments method to retrieve the external audience mapping for a specific user from the Optimizely Data Platform (ODP) server. The Feature Experimentation Python SDK lets you make the call to ODP in a synchronous or asynchronous fashion, depending on the callback parameter.
- If you do not provide a callback, the caller is blocked until the synchronous fetch is complete.
- Providing a callback does not block the caller.
fetch_qualified_segments is a method of the OptimizelyUserContext object. See OptimizelyUserContext for details.
Parameters
The following table describes the parameters for the fetch_qualified_segments method:
Parameter | Type | Description |
|---|---|---|
`options` (optional) | list[str] | A list of options for fetching qualified segments from ODP. |
`callback` (optional) | Callback function | A callback function to be called with the fetch result. |
Returns – synchronous call
If you do not provide a callback, the fetch_qualified_segments method returns True if the qualified segments array in the user context was updated.
Returns – asynchronous call
If you provide a callback, the fetch_qualified_segments method fetches the segments on a new thread and returns the thread handle.
- If the fetch completes, the Python SDK updates the qualified segments list in the user context and then calls the callback function with a success status.
- If the fetch fails, the SDK calls the callback function with a failure status.
If the Python SDK does not find an ODP audience in the datafile, it does return an empty qualified segments array without sending a request to the ODP server.
Note You can read and write to the qualified segments array instead of calling fetch_qualified_segments using get_qualified_segments and set_qualified_segments.
This lets you bypass the remote fetching process from ODP or use your own fetching service. This can be helpful when testing or debugging.
Example fetch_qualified_segments call
attributes = { "app_version": "1.3.2" }
user = optimizely_client.create_user_context("fs-id-123", attributes)
def fetch_callback(fetch_successful):
print(fetch_successful)
if fetch_successful:
decision = user.decide("flag1")
user.track_event("purchase_event")
# spawned thread is returned
fetch_thread: threading.Thread = user.fetch_qualified_segments(fetch_callback)
# thread must eventually be joined to ensure callback is run to completion
fetch_thread.join()attributes = { "app_version": "1.3.2" }
user = optimizely.create_user_context("user123", attributes)
# Without segment option
success = user.fetch_qualified_segments()
# With segment options
odp_segment_options = [OptimizelyOdpOption.IGNORE_CACHE, OptimizelyOdpOption.RESET_CACHE]
success = user.fetch_qualified_segments(options=odp_segment_options)
if success:
decision = user.decide("flag1")
user.track_event("myevent")The following diagram shows the network calls between your application, the Feature Experimentation Python SDK, and the ODP server when calling fetch_qualified_segments:
- Your application calls the fetch_qualified_segments method.
- The Feature Experimentation Python SDK makes a GraphQL call to ODP to fetch segments the user ID qualifies for.
- ODP responds with the qualified audience segments.
- The Python SDK automatically caches the fetched segments mapping the user ID to the qualified segments.
- The fetch_qualified_segments method returns a boolean if the qualified segments array was updated.
- Call the decide method to get the flag decision for a given user, a decision object that contains the flag's status (enabled or disabled), and the assigned variation key.
After you fetch the segments, they are cached. This means that if you request the segments again (when new user contexts are created), the Python SDK can retrieve the audience segment information from the cache rather than from the remote ODP server.
The cache is used for the fetch_qualified_segments call. This method is called on the user context (the user context is fixed, including the real-time audience the user qualifies for).
The cache only applies when calling the fetch_qualified_segments call. If you set the cache timeout to 0, the cache is disabled. Optimizely uses the least recently used (LRU) algorithm, so the oldest record is bumped out when the maximum size is reached. If there is a cache miss upon the method call, Optimizely makes a network request.
If you would like to bypass caching, you can add the following options to your OptimizelyOdpOption array:
- IGNORE_CACHE – Bypass the segments cache for lookup and save.
- RESET_CACHE – Reset the segments cache.
is_qualified_for
Minimum SDK version
5.0.0
Description
Check if the user is qualified for the given audience segment.
Parameters
The following table describes the parameters for the is_qualified_for method:
Parameter | Type | Description |
|---|---|---|
segment | String | The ODP audience segment name to check if the user is qualified for. |
Returns
True if the user is qualified.
Examples
The following is an example of whether the user is qualified for an ODP segment:
attributes = {"laptop_os": "mac"}
user = optimizely_client.create_user_context("fs_id_12", attributes)
success = user.fetch_qualified_segments()
qualified = user.is_qualified_for("segment1")Source files
The language and platform source files containing the implementation for Python are available on GitHub.