Microprediction is a leading open platform to share and score time series models. Microprediction can
- Clean and enrich live data
- Alert you to outliers and anomalies
- Provide you short term forecasts
- Identify patterns in model residuals
We work together with Microprediction to offer Amphora Data as an inbuilt data storage and collaboration tool.
To get started using Amphora and Microprediction together you need to install the appropriate SDKs with
You will also need an Amphora Data username and password (register for free here) as well as a key for Microprediction
You can use Amphora Data to store daily snap shots of all your agents and to share and use others pandemic data. This saves your computational cost, and gives you more data to train models on and research the dynamics of the system.
The latest code for using Amphora with Pandemic is on GitHub
Use the Amphora callback with
You will need to store your Amphora username and password as environment variables as well as the
amphora id if you want to push your data to an existing Amphora
Full code for using Amphora with Microprediction is below
We are looking to offer broader support and integration with Microprediction. Please come back later for more details or let us know which packages are best to support