data_gen_asym: Data Generation Asymptotic

In stressor: Algorithms for Testing Models under Stress

Data Generation Asymptotic

Description

Creates a synthetic data set for an additive asymptotic model. See the details section for clarification.

Usage

data_gen_asym(
  n,
  weight_mat = matrix(rlnorm(10), nrow = 2, ncol = 5),
  y_int = 0,
  resp_sd = 1,
  window = 1e-05,
  ...
)

Arguments

n	The number of observations for each parameter.
weight_mat	The parameter coefficients, where each column represents the coefficients and is two rows as each additive equation contains two parameters. Defaulted to be 10 random numbers from the log-normal distribution. The second row of the matrix needs to be positive.
y_int	The y-intercept term of the additive model.
resp_sd	The standard deviation of the epsilon term to be added for noise.
window	Used to determine for any given X variable to get you within distance to capture the asymptotic behavior.
...	Additional arguments that are not currently implemented.

Details

Observations are generated from the following model:

y = \sum_{i = 1}^n -\alpha_ie^{-\beta_i \cdot x_i} + y_{int}

Where 'n' is the number of parameters to be used, \alpha_i's are the scaling parameter and the \beta_i's are the weights associated with each x_i. With the y_{int} being where it crosses the y-axis.

Value

A data.frame object with the n rows and the response variable with the number of parameters being equal to the number of columns from the weight matrix.

Examples

 # Generates 10 observations
 asym_data <- data_gen_asym(10)
 asym_data

stressor documentation built on May 29, 2024, 3:09 a.m.