search

cjerzak / fastrerandomize-software

FastRerandomize: An R Package for Ultra-fast Rerandomization Using Accelerated Computing
https://arxiv.org/abs/2310.00861
MIT License
6 stars 0 forks source link

Create VIgnettes for Package #19

Open WarrenZhu050413 opened 2 weeks ago

WarrenZhu050413 commented 2 weeks ago

Similar to updating package tutorial #12, a vignette should be created for the function. This would be longer form for the tutorial. I think "The general analysis workflow must be illustrated by an enlighting non-trivial case study" (from jstatsoft is a good guideline for the vignette.

WarrenZhu050413 commented 1 week ago

In addition (and similar to) Vignettes, also tutorials:

Ideally 3 tutorials:

Similar to the agriculture tutorial we have now

Going through a similar synthetic experiment with larger n, and go through the monte_carlo rerandomization scheme to demonstrate the need for both acceleration and the use of monte algorithms. Here since the matrices are larger, we could introduce the approximate matrix inversion too.

Go through a synthetic experiment with larger dimensions, and show that the need for lower acceptance thresholds and herefore higher number of rerandomization increases with the dimensions