Learn R Programming

⚠️There's a newer version (1.1.0) of this package.Take me there.

An example for Shiny App usage

Taking the 2PL as an example, we illustrate how to use the Shiny app below.

In the first step, the server end (e.g., test administer, school board) can be launched by running the Shiny app (runserver()) with the interface shown below:

Then, the client-end Shiny app can be initialized (runclient()).

When the client first launches, it will automatically connect to the localhost port 8000 as default.

If the server is deployed on another computer, type the server's IP address and port (which will be displayed on the server's interface), then click "reconnect". The screenshots of the user interface are shown below.

Then, the client should choose a file to upload to the local Shiny app to do local calculations, without sending it to the server. The file should be a csv file, with either binary or graded response, and all clients should share the same number of items, and the same maximum score in each item (if the answers are polytomous), otherwise, there will be an error message suggesting to check the datasets of all clients.

After all the clients upload their data, the server should click "start" to begin the federated estimates process and after the model converges, the client should click "receive result". The server will display all item parameters and the client will display all item parameters and individual ability estimates.

The clients will also display bar plots of the ability estimates.

Copy Link

Version

Install

install.packages('FedIRT')

Monthly Downloads

186

Version

0.1.0

License

MIT + file LICENSE

Maintainer

Biying Zhou

Last Published

April 10th, 2024

Functions in FedIRT (0.1.0)

runserver

Server for Federated IRT Model Estimation
example_data_graded_and_binary

Graded Response Dataset for Federated Graded Model
fedirt_gpcm

Federated Graded Response Model Estimation Function
fedirt_2PL_data

Federated 2PL model
fedirt_gpcm_data

Federated gpcm model
fedirt_2PL_median_data

Federated 2PL model
fedirt_2PL

Federated 2PL estimate function
runclient

Client for Federated IRT Model Estimation
memoize

Memoization Function for Speed Optimization
g_logL_gpcm

Gradient of Log-Likelihood for the federated graded Model
logL

Log-Likelihood of the federated 2PL Model
logL_entry

Aggregate Log-Likelihood Function for Federated Learning
logL_gpcm

Log-Likelihood of the federated graded Model
g_logL

Gradient of Log-Likelihood for the federated 2PL Model
example_data_2PL_2

Binary Response Dataset for Federated 2PL Model
example_data_2PL

Binary Response Dataset for Federated 2PL Model
example_data_graded

Graded Response Dataset for Federated Graded Model
example_data_2PL_1

Binary Response Dataset for Federated 2PL Model
g_logL_entry

Aggregated Gradient of Log-Likelihood for Federated Learning