Skip to main content

Overview

  • A Data Set is a bundle of one or more CSV/JSON data and optional indexes and a manifest file describing the data as well as instructions which scope to install them in.
  • Couchbase offers a set of curated Data Sets in the Public Store (more on this in the next section).
  • You can create your own Data Set and store it in:
    • Your Local Store: This is on your Mac and private to you
    • A shared Cloud Store: To colloborate with your team

Example Data Set​

Full Local Repository​

~/Documents/MyRepository
├── Backups
│   ├── travel-sample
│   │   ├── 2023-10-05T09_43_34.506946-07_00
│   │   ├── 22023-10-24T09_48_26.419267-07_00
│   │   ├── 2024-01-20T12_39_35.08943-08_00
│   │	│	├── cluster-config.json
│   │	│	├── plan.json
│   │	│	├── travel-sample-090c5855f34364b6eaf63d18675fb1d7
│   │	│	│	├── gsi.json
│   │	│	│	├── gsi.metadata.json
│   │	│	│	├── data
│   │	│	│	│	├── *.*
│   │	│	│	├── ranges
│   │	│	│	│	├── *.*
│   │   ├── README.md
│   │   ├── backup-meta.json
│   ├── views
│   ├── model
│   ├── index.js
├── Datasets
│   ├── music
│   │   ├── couchmusic2-countries.json.zip
│   │   ├── couchmusic2-playlists-1.json.zip
│   │   ├── couchmusic2-playlists-2.json.zip
│   │   ├── couchmusic2-subregions.json.zip
│   │   ├── couchmusic2-tracks.json.zip
│   │   ├── couchmusic2-userprofiles.json.zip
│   │   ├── manifest.json
│   ├── movies
├── Favorites
│   ├── travel-sample
│   │   ├── inventory
│   │   │	├── 1.sql
│   │   │	├── 2.sql

The Datasets section​

  • Folder: Each Dataset must be in its own folder. The name is not displayed
  • Files: These can be json/csv/tsv files. They must all be of the same type. They must be zipped.
  • You must have a manifest.json file.

manifest.json​

Here is an example of the file for the music dataset

{
  "name": "Couch Music",
  "title": "Music",
  "subtitle": "Small. 5 Collections. 100 MB Size. Uploaded: June 11th, 2023",
  "description": "This Data Set is installed in **Scope: music** and has **5 Collections**: --countries: 258 docs-- --subregions: 23 docs-- --userprofiles: 49,981 docs-- --tracks: 97,216 docs-- --playlists: 132,844 docs-- with 14 GSI **indexes**. <br>*This is a good dataset if you are coming from a RDBMS background with multiple tables and JOIN operations*",
  "scope": "music",
  "fileType": "json",
  "formatType": "list",
  "files": [
    {
      "collection": "countries",
      "key": "country::%countryCode%",
      "docs": "couchmusic2-countries.json.zip",
      "indexes": [
        "CREATE INDEX country_code ON countries (countryCode)",
        "CREATE INDEX country_region_number ON countries (`region-number`)"
      ]
    },
    {
      "collection": "subregions",
      "key": "sub-region::%region-number%",
      "docs": "couchmusic2-subregions.json.zip",
      "indexes": ["CREATE INDEX region ON subregions (region)"]
    },
    {
      "collection": "userprofiles",
      "key": "userprofile::%username%",
      "docs": "couchmusic2-userprofiles.json.zip",
      "indexes": [
        "CREATE INDEX userprofile_address_country_code ON userprofiles (countryCode)",
        "CREATE INDEX userprofile_created ON userprofiles (created)",
        "CREATE INDEX userprofile_status ON userprofiles (status)",
        "CREATE INDEX userprofile_gender ON userprofiles (gender)",
        "CREATE INDEX userprofile_last_name ON userprofiles (lastName)",
        "CREATE INDEX userprofile_first_name ON userprofiles (firstName)"
      ]
    },
    {
      "collection": "tracks",
      "key": "track::%id%",
      "docs": "couchmusic2-tracks.json.zip",
      "indexes": [
        "CREATE INDEX track_artist ON tracks (artist)",
        "CREATE INDEX track_genre ON tracks (genre)",
        "CREATE INDEX track_title ON tracks (title)",
        "CREATE INDEX track_avg_rating ON tracks (avg_rating)"
      ]
    },
    {
      "collection": "playlists",
      "key": "playlist::%id%",
      "docs": "couchmusic2-playlists-1.json.zip"
    },
    {
      "collection": "playlists",
      "key": "playlist::%id%",
      "docs": "couchmusic2-playlists-2.json.zip"
    },
    {
      "collection": "playlists",
      "key": "playlist::%id%",
      "docs": "couchmusic2-playlists-3.json.zip",
      "indexes": ["CREATE INDEX playlist_owner ON playlists (owner.username)"]
    }
  ]
}

Please take a look at the details to be provided. This will display on the UI like so:

datasets-sample-music
  • Clicking this will import this Dataset into your Active Database
  • It will automatically create the required Scope and Collections
  • It will automatically create and build the indexes