gd_new: grad_descent() function

In Yannuo10/bis557: Homework 1

Description

another way to build a linear model. gradient descent for ordinary least squares

Usage

gd_new(form, data, contrasts = NULL, itr = 1e+05, L = 1e-04, tolerance = 1e-10)

Arguments

form	a formula;
data	a data frame used for the function;
contrasts	a list of contrasts for factor variables
itr	number of iteration
L	learning rate for gradient descent
tolerance	allowed tolerance for two residuals

Examples

data(iris)
fit_linear_model <- grad_descent(Sepal.Length ~ ., iris, contrasts = list(Species = "contr.sum"))

Yannuo10/bis557 documentation built on Dec. 21, 2020, 10:02 p.m.