The Timescale NFT Starter Kit is a step-by-step guide to get up and running with collecting, storing, analyzing and visualizing NFT data from OpenSea, using PostgreSQL and TimescaleDB.

The NFT Starter Kit will give you a foundation for analyzing NFT trends so that you can bring some data to your purchasing decisions, or just learn about the NFT space from a data-driven perspective. It also serves as a solid foundation for your more complex NFT analysis projects in the future.

We recommend following along with the NFT Starter Kit tutorial to get familiar with the contents of this repository.

For more information about the NFT Starter Kit, see the announcement blog post.

We provide multiple standalone components to help your data exploration journey at each level.

• Relational schema for storing NFT sales, assets, collections, and accounts.

• Data ingestion script, that collects historical data from OpenSea and ingests it into TimescaleDB. Read more!
• Sample data, that you can download and ingest to get started quickly. Read more!

• Streamlit dashboard, to analyze collection sales. Read more!
• Grafana dashboard template file
• Dockerized TimescaleDB + Apache Superset with pre-loaded data, to store and analyze NFTs.

• Sample queries to use as a starting point for your own analysis.

Whichever component you are most interested in, first clone the repository:

git clone https://github.com/timescale/nft-starter-kit.git
cd nft-starter-kit

This part of the project is fully Dockerized. TimescaleDB and the Superset dashboard is built out automatically using docker-compose. After completing the steps below, you will have a local TimescaleDB and Superset instance running in containers - containing 500K+ NFT transactions from OpenSea.

The Docker service uses port 8088 (for Superset) and 6543 (for TimescaleDB) so make sure there's no other services using those ports before starting the installation process.

• Docker

• Docker compose

  Verify that both are installed:

  docker --version && docker-compose --version

1. Run docker-compose up --build in the /pre-built-dashboards folder:

   cd pre-built-dashboards
   docker-compose up --build

   See when the process is done (it could take a couple of minutes):

   timescaledb_1      | PostgreSQL init process complete; ready for start up.

2. Go to http://0.0.0.0:8088/ in your browser and login with these credentials:

   user: admin
   password: admin

3. Open the Databases page inside Superset (http://0.0.0.0:8088/databaseview/list/). You will see exactly one item there called NFT Starter Kit.

4. Go check out your NFT dashboards!

   Collections dashboard: http://0.0.0.0:8088/superset/dashboard/1

   Assets dashboard: http://0.0.0.0:8088/superset/dashboard/2

If you'd like to ingest data into your database (be it a local TimescaleDB, or in Timescale Cloud) straight from the OpenSea API, follow these steps to configure the ingestion script:

• Python 3
• TimescaleDB installed
• Schema has been set up using the schema.sql script.

1. Go to the root folder of the project:

   cd nft-starter-kit

2. Create a new Python virtual environment and install the requirements:

   virtualenv env && source env/bin/activate
   pip install -r requirements.txt

3. Replace the parameters in the config.py file:

   DB_NAME="tsdb"
   HOST="YOUR_HOST_URL"
   USER="tsdbadmin"
   PASS="YOUR_PASSWORD_HERE"
   PORT="PORT_NUMBER"
   OPENSEA_START_DATE="2021-10-01T00:00:00" # example start date (UTC)
   OPENSEA_END_DATE="2021-10-06T23:59:59" # example end date (UTC)
   OPENSEA_APIKEY="YOUR_OPENSEA_APIKEY" # need to request from OpenSea's docs

4. Run the Python script:

   python opensea_ingest.py

   This will start ingesting data in batches, ~300 rows at a time:

   Start ingesting data between 2021-10-01 00:00:00+00:00 and 2021-10-06 23:59:59+00:00
   ---
   Fetching transactions from OpenSea...
   Data loaded into temp table!
   Data ingested!
   Data has been backfilled until this time: 2021-10-06 23:51:31.140126+00:00
   ---

   You can stop the ingesting process anytime (Ctrl+C), otherwise the script will run until all the transactions have been ingested from the given time period.

If you don't want to spend time waiting until a decent amount of data is ingested, you can just use our sample dataset which contains 500K+ sale transactions from OpenSea (this sample was used for the Superset dashboard as well)

• TimescaleDB installed
• PSQL (installation guide)

1. Go to the folder with the sample CSV files (or you can also download them from here):

   cd pre-built-dashboards/database/data

2. Connect to your database with PSQL:

   psql -x "postgres://host:port/tsdb?sslmode=require"

   If you're using Timescale Cloud, the instructions under How to Connect provide a customized command to run to connect directly to your database.
3. Import the CSV files in this order (it can take a few minutes in total):

   \copy accounts FROM 001_accounts.csv CSV HEADER;
   \copy collections FROM 002_collections.csv CSV HEADER;
   \copy assets FROM 003_assets.csv CSV HEADER;
   \copy nft_sales FROM 004_nft_sales.csv CSV HEADER;

4. Try running some queries on your database:

   SELECT count(*), MIN(time) AS min_date, MAX(time) AS max_date FROM nft_sales