mv_pn_test: Runs a multivariate point null test. This function returns an...

In amp: Statistical Test for the Multivariate Point Null Hypotheses

Runs a multivariate point null test. This function returns an approximate p-value for the specified test statistic.

Description

Runs a multivariate point null test. This function returns an approximate p-value for the specified test statistic.

Usage

mv_pn_test(obs_data, param_est = NULL, control = test.control())

Arguments

obs_data	The observed data to be used for finding the optimal norm (training), and finding the test statistic (testing). Similar to above, each row is an observation and each column corresponds to either the outcome (first column) or a covariate.
param_est	Function used to estimate the parameter and corresponding influence curve.
control	List used to define controls for test.

Value

The test will always return the following output:

- pvalue: The approximate value of the test statistic

- test_stat: The approximate value of the test statistic

- test_st_eld: The approximate limiting distribution of the test statistic (with length equal to ts_ld_bs_samp).

- chosen_norm: A vector indicating which norm was chosen by the adaptive test

- param_ests: The parameter estimate.

- param_ses: An estimate of the standard error of each element of param_ests

- oth_ic_inf: Any other information provided by the param_est function when calculating the IC and parameter estimates. Additional information may be returned by specifying it in the test.control function:

- If "var_est" is contained in other_output, the test output will contain will have var_mat returned which is the empirical second moment of the IC (equal asymptotically to the variance estimator).

- If "obs_data" is contained in the other_output, the test output will return the data passed to the testing function.

Examples

set.seed(10)
## NOTE: More monte-carlo samples should be taken are taken here.  This is
## only done to lower computation time.
test <- mv_pn_test(data.frame(y = rnorm(100), x = rnorm(100)),
                   ic.pearson, test.control(n_peld_mc_samples = 20,
                                            ts_ld_bs_samp = 20))

amp documentation built on April 6, 2022, 9:06 a.m.