With data in the Exchange, a data scientist, researcher, or developer can analyze it. The goal is usually a data product, like a dashboard, that a clinician or patient can use without writing code.
How it works¶
Work on the Hub. The Hub is a JupyterLab environment that logs you in with your Exchange account. The client library is already installed and set up to reach the Exchange. See the Hub documentation.
Or work anywhere else. The client library works in any Python environment, and other FHIR tools can read from the Exchange’s FHIR API.
Pull data into a DataFrame. The client library returns observations as pandas DataFrames, so you can explore them with the usual Python tools.
Turn a notebook into a data product. Voilà shows a notebook as a dashboard, without the code. Jupyter Book turns notebooks into reports and websites. The SMART on FHIR provider template turns a notebook into an app that clinicians open from their EHR (see step 4).
The demos are worked examples. Most pair a notebook for researchers with a dashboard for clinicians. The demos only run on the demo deployment at UC Berkeley.
Learn more¶
Explore your data: a first notebook, from a blank page.
Blood pressure tutorial: a getting started tutorial for pulling and plotting blood pressure data from the Exchange.
CGM tutorial: an end-to-end tutorial for pulling and analyzing continuous glucose monitor data from the Exchange.
Next step: Share data products with clinicians and patients.