gsd: Spatial dependent attribute with unknown skewness

In Sewanou/genspatattr: Generates skewed spatial attributes

Description

Generates a spatial dependent attributes of which skewness is provided as output.

Usage

gsd (size, grid, sk = 1, dep = 0.5, mu = 0, v = 1,  method = "SAR", dst = "skewnormal")

Arguments

size	a vector of two elements specifying respectively numbers of indivuals and time points of the attribute to be generated.
grid	number of rows and columns for grids.
sk	a scalar that represents the skewness coefficient to be considered.
dep	a scalar that represents the dependence correlation between values within a given attribute.
mu	the mean value
v	the variance value
method	a character string naming the the approach used to create spatial dependence between variables. The "SAR" method is the Kapoor, Kelejian, and Prucha (2007)-spatial autoregressive random effects (SAR-RE) model. The "SARMA" method represents the Lee and Yu (2012) spatial autoregressive moving average random effects (SARMA-RE) model.
dst	a character string naming the distribution to be used for vectors generation. "skewnormal" the skew-normal distribution is limited to a range of [0,1[ for the coefficient of skewness. "lognormal" the lognormal distribution allows a large range of coefficients skewness values.

Value

Returns a list:

spdata	the generated attribute.
skew.coef	effective skewness coefficient from generated data

Author(s)

Sewanou Honfo <[e-mail]> and Brezesky Kotanmi <[e-mail]> and Eunice Gnanvi <[e-mail]> and Chenangnon Tovissode <[e-mail]> and Romain glèlè Kakaï <[e-mail]> / LABEF_08_2019

References

Mohadeseh Alsadat Farzammehr, Mohammad Reza Zadkarami & Geoffrey J. McLachlan (2019): Skew-normal generalized spatial panel data model, Communications in Statistics - Simulation and Computation, DOI: 10.1080/03610918.2019.1622718

Sewanou/genspatattr documentation built on Oct. 30, 2019, 11:52 p.m.