Skip to main content
Version: 1.2.0

Dashboard API

Dashboard provides a short summary about the usage of Feature store.

Recently used projects​

To get overview about recently used projects, to list their names, descriptions, access times and optionally to get access to a project itself, use following methods:

recently_used_projects = client.dashboard.get_recently_used_projects()
recently_used_project = recently_used_projects[0]
recently_used_project.name
recently_used_project.description
recently_used_project.updated_at
recently_used_project.last_access_at
project = recently_used_project.get_project()

Recently used feature sets​

Similarly, to get overview about recently used feature sets, to list their names, descriptions, access times and to get access to a feature set itself, use following methods:

recently_used_feature_sets = client.dashboard.get_recently_used_feature_sets()
recently_used_feature_set = recently_used_feature_sets[0]
recently_used_feature_set.name
recently_used_feature_set.description
recently_used_feature_set.updated_at
recently_used_feature_set.last_access_at
feature_set = recently_used_feature_set.get_feature_set()

Feature sets popularity​

To provide hints about feature sets usage, Feature store tracks how often are individual feature sets retrieved (among all users) and provides a sorted list of those feature sets. A user can find out a feature set name, description, how many times was retrieved and its own access rights to that feature set.

feature_sets_popularity = client.dashboard.get_feature_sets_popularity()
popular_feature_set = feature_sets_popularity[0]
popular_feature_set.name
popular_feature_set.description
popular_feature_set.current_permission
popular_feature_set.number_of_retrievals
feature_set = popular_feature_set.get_feature_set()

Making list of favorite feature sets​

To simplify navigation across different feature sets, a user can mark a feature set to include it into a list of personal favorite feature sets. Whenever a feature set gets pinned then its reference is put onto the top of the list (it applies for a feature set pinned in the past as well). When a feature set is no more of interest it can be unpinned to remove it from the list. To get the favorite feature sets list, use a method from dashboard API.

note

The list method does not return feature sets directly. Instead, it returns an iterator which obtains the feature sets lazily.

fs = project.feature_sets.get("training_fs")

# adding a feature set to favorite list
fs.pin()

# getting the list of favorite feature sets
favorites = client.dashboard.list_pinned_feature_sets()

# accessing returned element
favorite = next(favorites)
favorite.name
favorite.description
favorite.updated_at
favorite.pinned_at
favorite_fs = favorite.get_feature_set()

# removing a feature set from favorite list
fs.unpin()
Feedback