bitmapist: a powerful analytics library for Redis
This Python library makes it possible to implement real-time, highly scalable analytics that can answer following questions:
- Has user 123 been online today? This week? This month?
- Has user 123 performed action "X"?
- How many users have been active have this month? This hour?
- How many unique users have performed action "X" this week?
- How many % of users that were active last week are still active?
- How many % of users that were active last month are still active this month?
- What users performed action "X"?
This library is very easy to use and enables you to create your own reports easily.
Using Redis bitmaps you can store events for millions of users in a very little amount of memory (megabytes). You should be careful about using huge ids (e.g. 2^32 or bigger) as this could require larger amounts of memory.
Additionally bitmapist can generate cohort graphs that can do following:
- Cohort over user retention
- How many % of users that were active last [days, weeks, months] are still active?
- How many % of users that performed action X also performed action Y (and this over time)
- And a lot of other things!
If you want to read more about bitmaps please read following:
Can be installed very easily via:
$ pip install bitmapist
Setting things up:
from datetime import datetime, timedelta from bitmapist import setup_redis, delete_all_events, mark_event,\ MonthEvents, WeekEvents, DayEvents, HourEvents,\ BitOpAnd, BitOpOr now = datetime.utcnow() last_month = datetime.utcnow() - timedelta(days=30)
Mark user 123 as active and has played a song:
mark_event('active', 123) mark_event('song:played', 123)
Answer if user 123 has been active this month:
assert 123 in MonthEvents('active', now.year, now.month) assert 123 in MonthEvents('song:played', now.year, now.month) assert MonthEvents('active', now.year, now.month).has_events_marked() == True
How many users have been active this week?:
print len(WeekEvents('active', now.year, now.isocalendar()))
Iterate over all users active this week:
for uid in WeekEvents('active'): print uid
If you're interested in "current events", you can omit extra
arguments. Events will be populated with current time automatically.
For example, these two calls are equivalent:
MonthEvents('active') == MonthEvents('active', now.year, now.month)
Additionally, for the sake of uniformity, you can create an event from
any datetime object with a
from_date static method.
MonthEvents('active').from_date(now) == MonthEvents('active', now.year, now.month)
Get the list of these users (user ids):
print list(WeekEvents('active', now.year, now.isocalendar()))
There are special methods
next returning "sibling" events and
allowing you to walk through events in time without any sophisticated
delta method allows you to "jump" forward or backward for
more than one step. Uniform API allows you to use all types of base events
(from hour to year) with the same code.
current_month = MonthEvents() prev_month = current_month.prev() next_month = current_month.next() year_ago = current_month.delta(-12)
Every event object has
period_end methods to find a
time span of the event. This can be useful for caching values when the caching
of "events in future" is not desirable:
ev = MonthEvent('active', dt) if ev.period_end() < now: cache.set('active_users_<...>', len(ev))
As something new tracking hourly is disabled (to save memory!) To enable it as default do::
import bitmapist bitmapist.TRACK_HOURLY = True
Additionally you can supply an extra argument to
mark_event to bypass the default value::
mark_event('active', 123, track_hourly=False)
Sometimes data of the event makes little or no sense, for example,
to filter out your premium accounts, or in A/B testing. There is a
UniqueEvents model for this purpose. The model creates only one
Redis key and doesn't depend on the date.
You can combine unique events with other types of events.
A/B testing example:
active_today = DailyEvents('active') a = UniqueEvents('signup_form:classic') b = UniqueEvents('signup_form:new') print "Active users, signed up with classic form", len(active & a) print "Active users, signed up with new form", len(active & b)
Generic filter example
def premium_up(uid): # called when user promoted to premium ... mark_unique('premium', uid) def premium_down(uid): # called when user loses the premium status ... unmark_unique('premium', uid) active_today = DailyEvents('active') premium = UniqueEvents('premium') # Add extra Karma for all premium users active today, # just because today is a special day for uid in premium & active_today: add_extra_karma(uid)
To get the best of two worlds you can mark unique event and regular bitmapist events at the same time.
def premium_up(uid): # called when user promoted to premium ... mark_event('premium', uid, track_unique=True)
Perform bit operations
How many users that have been active last month are still active this month?
active_2_months = BitOpAnd( MonthEvents('active', last_month.year, last_month.month), MonthEvents('active', now.year, now.month) ) print len(active_2_months) # Is 123 active for 2 months? assert 123 in active_2_months
Alternatively, you can use standard Python syntax for bitwise operations.
last_month_event = MonthEvents('active', last_month.year, last_month.month) this_month_event = MonthEvents('active', now.year, now.month) active_two_months = last_month_event & this_month_event
Work with nested bit operations (imagine what you can do with this ;-))!
active_2_months = BitOpAnd( BitOpAnd( MonthEvents('active', last_month.year, last_month.month), MonthEvents('active', now.year, now.month) ), MonthEvents('active', now.year, now.month) ) print len(active_2_months) assert 123 in active_2_months # Delete the temporary AND operation active_2_months.delete()
If you want to permanently remove marked events for any time period you can use the
last_month_event = MonthEvents('active', last_month.year, last_month.month) last_month_event.delete()
If you want to remove all bitmapist events use:
When using Bit Operations (ie
BitOpAnd) you can (and probably should) delete the results unless you want them cached. There are different ways to go about this:
active_2_months = BitOpAnd( MonthEvents('active', last_month.year, last_month.month), MonthEvents('active', now.year, now.month) ) # Delete the temporary AND operation active_2_months.delete() # delete all bit operations created in runtime up to this point bitmapist.delete_runtime_bitop_keys() # delete all bit operations (slow if you have many millions of keys in Redis) bitmapist.delete_temporary_bitop_keys()
With bitmapist cohort you can get a form and a table rendering of the data you keep in bitmapist. If this sounds confusing please look at Mixpanel.
Here's a simple example of how to generate a form and a rendering of the data you have inside bitmapist:
from bitmapist import cohort html_form = cohort.render_html_form( action_url='/_Cohort', selections1=[ ('Are Active', 'user:active'), ], selections2=[ ('Task completed', 'task:complete'), ] ) print html_form dates_data = cohort.get_dates_data(select1='user:active', select2='task:complete', time_group='days') html_data = cohort.render_html_data(dates_data, time_group='days') print html_data # All the arguments should come from the FORM element (html_form) # but to make things more clear I have filled them in directly
This will render something similar to this:
Copyright: 2012 by Doist Ltd.