my_grad: Gradient computation

In lucaaiello/myOpt: My Optimization Package

Description

This function provides a vectorized version of the gradient computation

Usage

my_grad(par, X, Y, verbose = T)

Arguments

par	Parameters values.
X	Data predictors.
Y	Response variables.
verbose	If set TRUE the function produce messages during the computation.

Value

Returns the vector containing the gradient scores.

Examples

# Having the parameters vector "par", the predictors matrix "X",
# and the outcome vector "Y",
# the gradient is calculated as follows:

set.seed(8675309)

# data simulation for example purposes

n = 1000

x1 = rnorm(n)
x2 = rnorm(n)
X <- cbind(rep(1,n),x1,x2)

Y = 1 + 0.5*x1 + 0.2*x2 + rnorm(n)

# random initial values for the parameters of the linear model

par <- rnorm(dim(X)[2])

# gradient calculation

my_grad(par, X, Y)

lucaaiello/myOpt documentation built on Dec. 21, 2021, 11:51 a.m.