my_grad_descent: Preform Linear Model with Gradient Descent Method

In tqchen07/bis557: Package Encapsulating work for BIS557

Description

This is a simple algorithm to solve a linear regression model.

Usage

my_grad_descent(
  form,
  data,
  contrasts = NULL,
  lambda = 1e-04,
  tolerence = 1e-15,
  beta_ini = 1,
  max_itr = 1e+06
)

Arguments

form	a formula object, e.g. y ~ x1 + x2
data	a data frame
contrasts	a list, whose entries are values (numeric matrices, functions or character strings naming functions) to be used as replacement values for the contrasts replacement function and whose names are the names of columns of data containing factors. If NULL, then contrasts are ommitted.
lambda	a numeric number indicating the learning rate for gradient descent algorithm
tolerence	a numeric number indicating the precision of the algorithm
beta_ini	a numeric number indicating the initial value for beta coeffficient
max_itr	an integer indicating the maximum number iterations

Examples

data(iris)
fit <- my_grad_descent(Sepal.Length ~ ., data = iris)
fit$coefficients

tqchen07/bis557 documentation built on Dec. 21, 2020, 3:06 a.m.