helpers function

Helper Functions

model_polytomous_3dindex creates indices extracting 3D stats

model_polytomous_3dresponse converts responses from 2D to 3D

hermite_gauss stores pre-computed hermite gaussian quadratures points and weights

nr_iteration updates the parameters using the Newton-Raphson method

model_polytomous_3dindex(u)

model_polytomous_3dresponse(u)

hermite_gauss(degree = c("20", "11", "7"))

nr_iteration(param, free, dv, h_max, lr, bound)

estimate_3pl_debug(tracking, k)

estimate_3pl_eval(true_params, t, a, b, c, t_free, a_free, b_free, c_free)

estimate_gpcm_debug(tracking, k)

estimate_gpcm_eval(true_params, n_c, t, a, b, d, t_free, a_free, b_free,
  d_free)

estimate_grm_debug(tracking, k)

estimate_grm_eval(true_params, n_c, t, a, b, t_free, a_free, b_free)

Arguments

• u: the observed response, 2d matrix, values start from 0
• degree: the degree of hermite-gauss quadrature: '20', '11', '7'
• param: the parameter being estimated
• free: TRUE to free parameters, otherwise fix parameters
• dv: the first and second derivatives
• h_max: the maximum value of h
• lr: the learning rate
• bound: the lower and upper bounds of the parameter
• tracking: estimation tracking information
• k: the number of iterations in estimation
• true_params: a list of true parameters
• t: estimated ability parameters
• a: estimated discrimination parameters
• b: estimated difficulty parameters
• c: estimated guessing parameters
• t_free: TRUE to estimate ability parameters, otherwise fix
• a_free: TRUE to estimate discrimination parameters, otherwise fix
• b_free: TRUE to estimate difficulty parameters, otherwise fix
• c_free: TRUE to estimate guessing parameters, otherwise fix