Legato: Ruby Client for the Google Analytics Core Reporting and Management API

Check out the Wiki!

Feel free to open an issue if you have a question that is not answered in the Wiki

Chat!

We're trying out chat using Gitter Chat. I'll try to be in there whenever I'm at my computer.

If you've come here from Garb, welcome! There are a few changes from Garb, so you'll want to check out:

• Model Data
• Query Parameters
• And the biggest difference: Filtering

If you're not able to upgrade quite yet, Garb has been maintained https://github.com/Sija/garb

Google Analytics Management

1. Get an OAuth2 Access Token from Google, Read about OAuth2

   access_token = OAuth2 Access Token # from Google
   

2. Create a New User with the Access Token

   user = Legato::User.new(access_token)
   

3. List the Accounts and Profiles of the first Account

   user.accounts
   user.accounts.first.profiles
   

4. List all the Profiles the User has Access to

   user.profiles
   

5. Get a Profile

   profile = user.profiles.first
   

6. The Profile Carries the User

   profile.user == user #=> true
   

7. The profile can also lookup its "parent" Web Property

   profile.web_property
   

Google Analytics Model

class Exit
  extend Legato::Model

  metrics :exits, :pageviews
  dimensions :page_path, :operating_system, :browser
end
profile.exit #=> returns a Legato::Query
profile.exit.each {} #=> any enumerable kicks off the request to GA

Metrics & Dimensions

http://code.google.com/apis/analytics/docs/gdata/dimsmets/dimsmets.html

metrics :exits, :pageviews
dimensions :page_path, :operating_system, :browser

Filtering

Create named filters to wrap query filters.

Here's what google has to say: http://code.google.com/apis/analytics/docs/gdata/v3/reference.html#filters

Examples

Inside of any

Legato::Model

filter

metrics

dimensions

Return entries with exits counts greater than or equal to 2000

filter(:high_exits) {gte(:exits, 2000)}

or ...
filter :high_exits, &lambda {gte(:exits, 2000)}

Return entries with pageview metric less than or equal to 200

filter(:low_pageviews) {lte(:pageviews, 200)}

Filters with dimensions

filter(:for_browser) {|browser| matches(:browser, browser)}

Filters with OR

filter(:browsers) {|*browsers| browsers.map {|browser| matches(:browser, browser)}}

Using and Chaining Filters

Pass the profile as the first or last parameter into any filter.

Exit.for_browser("Safari", profile)

Chain two filters.

Exit.high_exits.low_pageviews(profile)

Profile gets a method for each class extended by Legato::Model

Exit.results(profile) == profile.exit

We can chain off of that method, too.

profile.exit.high_exits.low_pageviews.by_pageviews

Chaining order doesn't matter. Profile can be given to any filter.

Exit.high_exits(profile).low_pageviews == Exit.low_pageviews(profile).high_exits

Be sure to pass the appropriate number of arguments matching the lambda for your filter.

For a filter defined like this:

filter(:browsers) {|*browsers| browsers.map {|browser| matches(:browser, browser)}}

We can use it like this, passing any number of arguments:

Exit.browsers("Firefox", "Safari", profile)

Google Analytics Supported Filtering Methods

Google Analytics supports a significant number of filtering options.

Here is what we can do currently: (the operator is a method available in filters for the appropriate metric or dimension)

Operators on metrics (method => GA equivalent):

eql     => '==',
not_eql => '!=',
gt      => '>',
gte     => '>=',
lt      => ' '<='

Operators on dimensions:

matches          => '==',
does_not_match   => '!=',
contains         => '=~',
does_not_contain => '!~',
substring        => '[email protected]',
not_substring    => '[email protected]'

Session-level Segments

Your query can have a session-level segment, which works with filter expressions. It works like an advanced segment, except you don't have to create it beforehand, you can just specify it at query time.

Some of the numbers you'll get will be different from using a filter, since the subset of visits matched happens before dimensions and metrics are calculated (hover on the

segment

Some metrics and dimensions are not allowed for segments, see the API documentation for more details.

Note: Legato does not support Users vs Sessions, yet. The default will be sessions (the equivalent of the earlier, now removed, dynamic segments).

Defining, using and chaining segments

Return entries with exits counts greater than or equal to 2000

segment :high_exits do
  gte(:exits, 2000)
end

Return entries with pageview metric less than or equal to 200

segment :low_pageviews do
  lte(:pageviews, 200)
end

You can chain them

Exit.high_exits.low_pageviews(profile)

and call them directly on the profile

profile.exit.high_exits.low_pageviews

Accounts, WebProperties, Profiles, and Goals

Legato::Management::Account.all(user)
Legato::Management::WebProperty.all(user)
Legato::Management::Profile.all(user)
Legato::Management::Goal.all(user)

Other Parameters Can be Passed to a call to #results

• :start_date - The date of the period you would like this report to start
• :end_date - The date to end, inclusive
• :limit - The maximum number of results to be returned
• :offset - The starting index
• :sort - metric/dimension to sort by
• :quota_user - any arbitrary string that uniquely identifies a user (40 characters max)
• :samplinglevel - 'FASTER' or 'HIGHERPRECISION' https://developers.google.com/analytics/devguides/reporting/core/v3/reference#samplingLevel
• :segment_id - this will supersede any segments chained to the query

Real Time Reporting

https://developers.google.com/analytics/devguides/reporting/realtime/v3/ https://developers.google.com/analytics/devguides/reporting/realtime/dimsmets/

GA provides an endpoint to do basic reporting in near-realtime. Please read the above documentation to know which features (and dimentsion/metrics) are or are not available. It is also only available in beta so you must already have access.

Inside of Legato, you can simply add

realtime

#results

Query

Exit.results(profile).realtime

The results you iterate over (with

.each

You can also call

realtime

Query

query = Exit.realtime
query.realtime? #=> true
query.tracking_scope #=> 'rt'

Managing Quotas

Assigning a

quota_user

user_ip

user = Legato::User.new(access_token)
user.quota_user = 'some_unique_user_identifier'
# OR
user.user_ip = ip_address_from_a_web_user_or_something