Models
The PMN allows users to directly access models on the cloud. Users can interact with models without worrying about setting up the required software environment and accessing high-performance computing power. Models can be accessed using Microsoft Excel, R, Python, JavaScript, or pretty much any other programming language that supports standard RESTful APIs.
Access Models Through Microsoft Excel
Many of the models hosted in the repository come with a macro-enabled Microsoft Excel file that simplifies access
to the models. The Excel spreadsheet includes a list of model inputs, which can be changed by the user, and
automatically populates numerical and graphical results. All you need to do is download the spreadsheet, change
input parameters to those of your choice, and click Run.
Access Models Through R
The client R package peermodels can be used to easily access the models hosted on the Peer Models
Network cloud. You can use the package to retrieve default input parameters for a model, change the parameters as
needed, and run the model remotely with the new parameters. You will need an API key to access the model of your
choice — read below to find out how to apply for one.
REST APIs
The PMN allows users to directly access models on the cloud using standard RESTful APIs that can be called from
Python, JavaScript, and pretty much any modern programming language that supports HTML APIs. Many of the models in
the repository include sample API calls in curl. You will need an API key to access the model of your
choice — read below to find out how to apply for one.
Get Your API Key
Our current funding from provincial and federal agencies enables us to provide free access to the models hosted on the cloud. If you need an API key to access a model of your choice on the PMN cloud, send us a quick note with your name, affiliation, use case, and the name of the model you would like to access.
Share Your Models
The current version of the PMN cloud is able to host decision-analytic and clinical prediction models developed in R. If you have a model that you’d like to share with others, contact us to learn how.