Surveyor Warehouse

Surveyor Warehouse allows you to extract and transform the responses from surveyor response sets into an alternate database table structure. This can be generally helpful for querying surveyor responses within a table structure of one's choice.

The alternate database table structure is defined by adding a :data_export_identifier attribute to each question with the format "table.column" (e.g. 'subjects.name').

Example

Given the survey:

  q 'What is your name?', :data_export_identifier => 'subjects.name'
  a :string

  q 'What is your age?', :data_export_identifier => 'subjects.age'
  a :integer

  q 'What is your gender?', :pick => :one, :data_export_identifier => 'subjects.gender'
  a 'Male', :reference_identifier => "male"
  a 'Female', :reference_identifier => "female"

  q 'Which brand of shoes do you wear?', :pick => :any, :data_export_identifier => 'subjects.shoes'
  a 'Nike', :reference_identifier => 'nike'
  a 'Adidas', :reference_identifier => 'adidas'
  a 'None', :omit, :reference_identifier => 'none'

  repeater 'List all your siblings' do
    q 'Name', :data_export_identifier => 'siblings.name' 
    a :string
  end

Responses would be transformed into:

  db=# select * from subjects;
    name   | age | gender |     shoes     | access_code |      id
  ---------+-----+--------+---------------+-------------+--------------
   raphael |  18 | male   | {nike,adidas} | QUhwHtMPyQ  | QUhwHtMPyQ.0
   stephen |  50 | male   | {none}        | zZu1OV1hmw  | zZu1OV1hmw.0
  (2 rows)

  db=# select * from siblings;
       name     | access_code |      id
  --------------+-------------+--------------
   leonardo     | QUhwHtMPyQ  | QUhwHtMPyQ.0
   michelangelo | QUhwHtMPyQ  | QUhwHtMPyQ.1
   elizabeth    | zZu1OV1hmw  | zZu1OV1hmw.0
   james        | zZu1OV1hmw  | zZu1OV1hmw.1
  (4 rows)

Getting Started

1. Survey questions must have data_export_identifiers with the format "table.column" (e.g. 'subjects.name')
2. Take the survey
3. bundle exec rake surveyor:warehouse
4. Query the results

Limitations

• Currently only works with PostgresSQL