Social network sample database — MongoDB, flat (390,729 rows)

A collection per table. Each primary key is the document _id; foreign keys are fields under their own names, so the SQL version's joins become $lookup stages.

Load it:
  sh import.sh                    (into mongodb://localhost:27017)
  sh import.sh "<connection string>"

Or one collection at a time: mongoimport --db social-network --collection <name> --file <name>.ndjson

Collections:
  blocks — 12,500 documents
  comment_likes — 15,000 documents
  comments — 20,000 documents
  conversation_participants — 12,500 documents
  conversations — 1,563 documents
  devices — 12,500 documents
  follows — 10,000 documents
  group_memberships — 12,500 documents
  group_posts — 12,500 documents
  groups — 1,563 documents
  hashtags — 40 documents
  login_history — 12,500 documents
  media — 3,000 documents
  message_receipts — 10,000 documents
  messages — 10,000 documents
  notifications — 12,500 documents
  page_admins — 12,500 documents
  page_likes — 12,500 documents
  page_posts — 12,500 documents
  pages — 1,563 documents
  post_hashtags — 12,500 documents
  post_likes — 100,000 documents
  post_media — 12,500 documents
  posts — 12,000 documents
  profiles — 12,500 documents
  reports — 12,500 documents
  stories — 1,000 documents
  story_views — 5,000 documents
  user_tags — 12,500 documents
  users — 12,500 documents

Types: numeric columns are Decimal128, so totals stay exact; dates and timestamps are
Date values read as UTC (the data has no time zone); times and UUIDs are strings;
JSON columns are embedded objects.

Synthetic data, CC0-1.0, from databasesample.com. The same rows as the SQL and CSV files.
