Create a dataset from scratch and publish it with Datahub Cloud

Yoana Popova — Fri, 09 Aug 2024 14:30:08 +0000

In our previous article, we talked about DataDatahub Cloud: your stupidly simple and fast tool for turning your data stories or datasets on GitHub into a published, shareable site. It converts raw data and Markdown files into beautifully presented, interactive sites.

Today, we're going to tell you how to publish a dataset (multiple data files or a single data file) with DataHub Cloud.

As an example we're going to use an example dataset with an analysis of the top 1000 global universities:
https://www.kaggle.com/datasets/zahrayazdani81/univercitiesranking?resource=download

What You'll Need

GitHub account and basic knowledge of GitHub UI (especially editing and adding files)
A DataHub Cloud account.

Step 1: Create a GitHub repository with the data files and README.md file

Any DataHub Cloud site is built off of a GitHub repository. This is where you'd put all your dataset file(s) and any related markdown content that you want to publish. For the sake of simplicity, in this tutorial, we're only going to use a single README.md file. It's going to serve as a landing page for our site.

Tip
Any README.md or index.md file, either in a root of the repository or in a subfolder, will be treated as a "landing" page (of the whole site or a given folder) by the DataHub Cloud.

Go to your GitHub account and create a new repository. Note, you can check "Add a README file" checkbox. This will make GitHub automatically add an empty README.md file to our repository.

Hint
If you're new to GitHub, here are simple instructions on creating a repository: https://docs.github.com/en/repositories/creating-and-managing-repositories/quickstart-for-repositories

Now, let's continue this tutorial in our docs: https://datahub.io/docs/Create+a+dataset+from+scratch+and+publish+it+with+Datahub+Cloud

You can check out how a published dataset page looks like here: https://datahub.io/core/co2-ppm
You can check out how a published data story page looks like here:
https://datahub.io/@cheredia19/us-cities-population

Datahub Cloud: a simple and fast tool to turn your data stories or datasets on GitHub into a published, shareable site

Yoana Popova — Wed, 07 Aug 2024 13:37:21 +0000

Datahub Cloud is your stupidly simple and fast tool for turning your data stories or datasets on GitHub into a published, shareable site. It converts raw data and Markdown files into beautifully presented, interactive sites.

NOTE:
The current version of Datahub Cloud runs only off Github.

How to deploy your first site within seconds

The quickest way to start publishing with Datahub Cloud is to publish our template and customize it to fit your needs. You can do that in 5 simple steps:

Click on the selected template from the ones listed below and click "Use this template" at the top right to create a new repository

TEMPLATES:

Data-rich story template that renders like this

Dataset template that renders like this

PKM template that renders like this

NOTE
If you don't have any data files or notes, you can also start from scratch (go to Start from scratch)

Go to the app and create a new site by selecting the repository you just created (leave the "Root Dir" field empty)

Done, it is now published! Just hit the green "Visit" button at the top right to see what it looks like.

❓ What to do after you publish your first site

After publishing our template and getting a feel for how it works, you have several options:

👷‍♂️ Customize your template
📈 Add visuals
🏖️ Publish another repo of your own
🆕 Create a new repo
🖋️ Publish your Obsidian notes

You can also take the template to the next level:

DEV Community: Yoana Popova

Create a dataset from scratch and publish it with Datahub Cloud

What You'll Need

Datahub Cloud: a simple and fast tool to turn your data stories or datasets on GitHub into a published, shareable site

How to deploy your first site within seconds

❓ What to do after you publish your first site