Contrast: Contrast estimated from an object inheriting from...

Description Methods

Description

An R6 class that stores results for a contrast of the form L θ - m estimated using a specified object inheriting from FamModelFit. Wald tests and confidence intervals for individual rows of the contrast matrix as well as an overall Wald chi-square test of the null hypothesis L θ - m = 0 are provided.

Methods

Public methods


Method new()

Constructs a new instance of this class.

Usage
Contrast$new(model_fit, L_mat, m)
Arguments
model_fit

An object inheriting from FamModelFit.

L_mat

A contrast vector (1 df) or matrix (>1 df) containing one contrast in each row. The contrast vector must have a number of elements equal to the number of model parameters. The contrast matrix must be of full row rank and have a number of columns equal the number of model parameters.

m

An optional vector containing the null value for each contrast. Will be set to the zero vector of length nrow(L_mat) if not specified.


Method get_model_fit()

Returns model_fit object.

Usage
Contrast$get_model_fit()

Method get_L_mat()

Returns L_mat.

Usage
Contrast$get_L_mat()

Method get_m()

Returns m.

Usage
Contrast$get_m()

Method get_L_theta_hat_m()

Returns L \hat{θ} - m.

Usage
Contrast$get_L_theta_hat_m()

Method get_V_L_theta_hat()

Returns L \hat{V}(\hat{θ}) L^{'}.

Usage
Contrast$get_V_L_theta_hat()

Method get_X2()

Returns Wald chi-square statistic for null hypothesis L θ - m = 0, which is (L \hat{θ} - m)^{'} [L \hat{V}(\hat{θ}) L^{'}]^{-1} (L \hat{θ} - m).

Usage
Contrast$get_X2()

Method get_df_X2()

Returns degrees of freedom of Wald chi-square statistic, which is the rank of L.

Usage
Contrast$get_df_X2()

Method get_p_X2()

Returns p-value for Wald chi-square statistic.

Usage
Contrast$get_p_X2()

Method print()

Formatted printing of the Contrast object.

Usage
Contrast$print(...)
Arguments
...

Arguments passed on to print_ests().


Method clone()

The objects of this class are cloneable with this method.

Usage
Contrast$clone(deep = FALSE)
Arguments
deep

Whether to make a deep clone.


kinnamon-lab/FamModel documentation built on April 26, 2021, 4:38 a.m.