Main concepts

Creating reports are simple. You add a few app sections, blocks and marks and then your run app.deploy()

from relevanceai import Client
client = Client(token=YOUR_AUTH_TOKEN)

from relevanceai.apps import ReportApp
app = ReportApp(name="My App", dataset=ds)
app.h1("My first app")
app.p("Creating my first app")
app.deploy()

You can also display your app within jupyter notebook with:

app.gui()

Blocks

Blocks are essentially content, every time you add a paragraph, image or chart that's a block. There are many different kind of blocks, some blocks can contain other blocks. A page is a combination of different blocks you add to it.

Marks

Marks are a way to format text blocks, such as bold, italic. You can add multiple marks to a single block.

Sections

Sections are a logical combination of blocks that is packaged together for a specific use case. e.g. you can have a combination of a dataframe of a time series and a line chart of that dataframe as a "time series section".

Deployable ID

Every app is associated with a deployable configuration. Every deployable has a unique id. You can access your app's id with app.deployable_id. If your id is None, run app.deploy and an id will be generated with the app.

You can reload apps by specifying its deployable id:

my_id = "123456"
app = ReportApp(name="My App", dataset=ds, deployable_id=my_id)

This will reload all the created blocks, marks and sections for the app into your app class.