pyfreya package

Subpackages

• pyfreya.cohort package
  • Submodules
  • pyfreya.cohort.cohort module
    • Short Tutorial in the Cohort Class.
      • Retention
      • Daily Active Users
      • Datetime
      • Revenue
        • Cohort Class
  • Module contents
• pyfreya.retention package
  • Submodules
  • pyfreya.retention.fit_functions module
  • pyfreya.retention.retention module
    • Short Tutorial in the Retention Class.
      • Custom Functions
        • Retention Class
  • Module contents
• pyfreya.revenue package
  • Submodules
  • pyfreya.revenue.revenue module
    • Short Tutorial in the Revenue Classes.
      • Revenue Classes
  • Module contents

Submodules

pyfreya.pyfreya module

Main module.

pyfreya.pyfreya.create_cohort(new_users, days_since_install=None, retention_values=None, retention_function='power', retention_profile=None, start_date=1, revenue_profile=None, name='')

Creates a cohort class. “new_users” parameter must be provided. Retention information must also be provided, by either: add retention values and days since install values or supply a pre-made retention profile - see Retention. A revenue profile can also be attached to a cohort - see Short Tutorial in the Revenue Classes..

The main variables of the class to keep track of is:

• df_user_dist: Contains information about the user by days since install (index of the pandas dataframe) and date (column of the pandas dataframe).
• df_dau: Contains information about daily active user and revenue. The index is date and the columns are dau, revenue and revenueUnc. Assuming each measure have been calculated.

Parameters:

• new_users – The amount of starting users.
• days_since_install (Optional[List[int]]) – The days since install values to go along with retention_values.
• retention_values (Optional[List[Union[float, Variable]]]) – The retention values to go along with days_since_install.
• retention_function – Function to fit the retention to.
• retention_profile (Optional[Retention]) – A premade retention profile using the Retention class.
• start_date (Union[datetime, date, int]) – The start date of the first cohort.
• revenue_profile (Optional[BaseRevenue]) – A revenue profile object who had inherited its behaviour after BaseRevenue.
• name – Name of cohort - is mostly used as identifier when working with multiple cohorts.

Returns:

pyfreya.pyfreya.create_retention(days_since_install, retention_values)

Creates a retention profile that can be used in cohorts.

Parameters:

• days_since_install (List[int]) – Days since install that accompanies the retention values.
• retention_values (List[Union[float, Variable]]) – Retention values can either be formatted in values below like

>>> [0.5, 0.05, 0.01]

or

>>> [50, 5, 1]

for values 50%, 5% and 1%. :return:

pyfreya.pyfreya.load_class(filename)

Parameters:filename (str) – Returns:

pyfreya.pyfreya.multi_cohort_dau_plot(cohorts, kind='line')

Plot DAU by date for each cohort. With kind=’line’ a line plot is used, where all values are plotted from 0. With kind=’bar’ a stacked bar plot is used, meaning each value on each date is placed on top of each other.

Parameters:

• cohorts (List[Cohort]) – List of cohotrs.
• kind – Type of plot, choose between line (default) and bar.

Returns:

pyfreya.pyfreya.multi_cohort_ret_plot(cohorts)

Plot retention for multiple cohorts.

Parameters:cohorts (List[Cohort]) – List of cohorts. Returns:

pyfreya.pyfreya.multi_cohort_rev_plot(cohorts, kind='line', cumulative=True)

Plot revenue from multiple cohorts. With kind=’line’ a line plot is used, where all revenue is plotted from 0. With kind=’bar’ a stacked bar plot is used, meaning each revenue on each date is placed on top of each other.

Parameters:

• cohorts (List[Cohort]) – List of cohorts.
• kind – Type of plot, choose between line (default) and bar.
• cumulative – Denotes if the cumulative lines should be included.

Returns:

pyfreya.pyfreya.save_class(filename, class_instance)

Parameters:

• filename (str) –
• objet –

Returns:

Module contents

Top-level package for PyFreya.