Create initial warehouse star schema

[blarghmatey]

User Story

• As an analytics engineer I would like to have a well-factored star schema to improve the flexibility and re-use of data assets in the warehouse

Description/Context

For detailed context refer to https://github.com/mitodl/hq/pull/5749

This epic covers the creation of the initial set of fact and dimension tables needed to support building the current "mart" tables from those underlying assets.

Acceptance Criteria

• All of the current "mart" tables can be created as views or materialized tables that are constructed from the fact and dimension tables
• All fact and dimension tables are fully documented such that any data consumer will understand the meaning of each field and record in each table

Plan/Design

Create the following set of dimensional tables (subject to adjustment as we progress through implementation):

Dimensions

• dim_content: Contains information about content objects including courses,
  course blocks, videos, site pages (e.g. in OCW), etc.
• dim_platform: Information about the technical system that a given event is
  executed within the context of.
• dim_user: Contains information about system users including
  demographic/profile attributes. This will need to be a
  slowly-changing-dimension model.
• dim_purchaseable: Includes information about elements of content and
  services that can be sold for >=$0 (e.g. programs, course runs, bootcamps,
  subscriptions, etc.)

Facts

• tfact_purchase: Monetary transactions (including $0 transactions, coupon
  redemption, financial aid, etc.)
• afact_grade: Information about a grade computed for a user in a course-run
• tfact_account: Contains events related to account actions (e.g. created,
  logged in, etc.)
• tfact_submit: Events involving a learner submitting information to a system,
  such as answering a question
• afact_credential: Information about the granting of a credential
  (e.g. course certificates, program certificates, etc.)
• tfact_coupon: Events pertaining to ecommerce coupons including creation,
  consumption, deletion/deactivating
• tfact_submit: Contains event details about information submitted by a user
  into a platform, such as answering a question, executing a search, posting a
  forum question, etc.
• tfact_video: Contains event information about interactions with a video such
  as play, pause, seek, etc.

Draft ERD (https://erdlab.page.link/7agMHFfKTiL2yxD57)

[KatelynGit]

https://erdlab.page.link/xjxGkH2X17N8shaR9

proposing a new model for engagement data to answer questions the xpro team has that aren't answerable in the current marts. Some questions they have are here: https://github.com/mitodl/hq/issues/6176.

Since a design couldn't be agreed on I'm not implementing this but I'll leave it here in case it's useful in the future