Both the R script "iterativeML" and the estimation.JS estimate the parameters the Rasch model/Bradley-Terry_Luce model with a conditional maximum likelihood procedure. This procedure is an iterative procedure using Newton's method of optimization. This procedure was initially developed by Rasch (1960) and implemented by Pollitt (2012) and NoMoreMarking ltd. among others.
Both Scripts were based on the code of NoMoreMarking ltd..
The estimation.js is tested in the estimation.test.js in conjunction with the "data for testing.R" file. The estimation is working good.
The estimateCJ and CML functions are the translation of the NoMoreMarking ltd.algorithms for use in the D-PAC tool. And the ConvertData function is to convert the data that the D-PAC tool provides to data the estimateCJ and CML functions can handle.
raschStats.js contains functions to calculate the Rasch probability and the Fischer Information.