The open model hub is a toolkit that contains functions and models that can be applied on data collected with Objectiv’s Tracker. The following types of functions/models are available:
- Helper functions: Simplify manipulating and analyzing the data.
- Aggregation models: Enable running some of the more common data analyses and product analytics metrics.
- Machine learning models: ML models such as logistic regression.
- Funnels: To analyze Funnels, e.g. discover all the (top) user journeys that lead to conversion or drop-off.
Helper functions always return a series with the same shape and index as the
DataFrame they are applied to. This ensures they can be added as a column to that
DataFrame. Helper functions that return
SeriesBoolean can be used to filter
the data. The helper functions can be accessed with the
map accessor from a
model hub instance.
|Labels all hits in a session True if that session is the first session of that user in the data.|
|Labels all hits True if the user is first seen in the period given |
|Labels a hit True if it is a conversion event, all other hits are labeled False.|
|Counts the total number of conversions given a partition (ie session_id or user_id).|
|Counts the number of time a user is converted at a moment in time given a partition (ie 'session_id' or 'user_id').|
|Returns a count backwards from the first conversion, given the partition.|
Aggregation models perform multiple Bach instructions that run some of the more common data analyses or
product analytics metrics. Always return aggregated data in some form from the
DataFrame the model is applied to. Aggregation models can be accessed with the
aggregate accessor from a model hub instance.
|Calculate the unique users in the Objectiv |
|Calculate the unique sessions in the Objectiv |
|Calculate the duration of sessions.|
|Calculate a frequency table for the number of users by number of sessions.|
|Finds the number of users in a given cohort who are active at a given time period, where time is computed with respect to the beginning of each cohort.|
|Calculate the top used features in the product.|
|Calculates what users did before converting by combining several models from the model hub.|
Machine learning models
Currently we support
LogisticRegression directly on Bach DataFrames
Currently we support
FunnelDiscovery directly on Bach DataFrames.