Get started with TriplyDB¶
This page takes you from nothing to a dataset you can query, in four steps. It should take about fifteen minutes.
Everything optional has been left out. Each step links to the full guide if you want the detail.
Before you start¶
You need an account on a TriplyDB instance — triplydb.com or your
organisation's own. Sign up or log in before going further. Step 2 loads a file
from triplydb.com, so an organisation's own instance needs to be able to reach
the internet.
You do not need to know SPARQL. Step 4 gives you a query to paste.
1. Create a dataset¶
From the home screen, open your account through "Your account" under "Quick links". On the account page, click the New button in the top-right corner and choose Dataset.
Enter iris as the dataset name. Everything else in the dialog is optional
and can be filled in later.
New datasets are Private, which is the right setting while you are experimenting. Nobody else can see it.
Click Add dataset. The dataset is created and its Graphs page opens.
Full guide: Add data.
2. Add data¶
This tutorial loads the Iris flower dataset that Triply publishes on triplydb.com: measurements of 150 iris flowers, converted to linked data. It is an 18 KB file, so it loads in seconds.
On the Graphs page, click Import from URL and paste this address into the URL field:
https://api.triplydb.com/datasets/Triply/iris/download.trig.gzClick Import. When the import finishes, the Graphs page lists five graphs holding 1,749 statements between them.
Files on your computer and graphs from other datasets can be loaded in the same place; Add data covers those routes.
Your data now lives in one or more graphs inside the dataset. See TriplyDB at a glance.
3. Look at what you loaded¶
Two views are available immediately, with no further setup:
- The linked data browser shows one resource at a time and lets you follow properties to the next one.
- The linked data table shows every triple, with columns for subject, predicate, object and graph.
Full guide: Data views.
4. Run a SPARQL query¶
Click SPARQL in the left-hand sidebar of your dataset. The query runs on Speedy, TriplyDB's built-in SPARQL engine, so there is nothing to set up first.
Paste this in:
select ?subject ?predicate ?object
where {
?subject ?predicate ?object.
}
limit 10Click Execute query. Ten triples come back — proof that the dataset and the query editor are both working.
That query matches everything, which is why it always returns something and why it is never useful beyond this moment. To write queries that answer real questions, start with SPARQL: the basics.
What next¶
| If you want to | Go to |
|---|---|
| Share the dataset with others | Publish data |
| Understand what a graph or service is | TriplyDB at a glance |
| Learn to write queries | SPARQL: the basics |
| Get the data back out | Export data |
| Automate loading it | Build your first flow |