ols_gd_hw2b: OLS with Gradient Descent

In brian-d1018/bis557: Brian's BIS 557 Work

Description

This function uses the gradient descent algorithm (matrix form) to solve OLS using 10-fold cross validation (to calculate out-of-sample accuracy).

Usage

ols_gd_hw2b(form, d, b_0, learn_rate, max_iter, contrasts = NULL)

Arguments

form	linear model formula
d	data frame
b_0	parameters of model: column vector of initialized coefficients
learn_rate	the initialized learning rate (aka step size)
max_iter	the maximum number of iterations for this algorithm
contrasts	optional list of constants for factor variables aka contrast

Value

the estimated coefficients, as well as the penalty (loss)

Examples

set.seed(2020)
x <- -30:30
df <- data.frame(y = 6 + 7*x + rnorm(length(x)), x = x)
bis557::ols_gd_hw2b(form = y ~ ., d = df, b_0 = rep(1e-16, 2),
                    learn_rate = 0.18, max_iter = 1e5)

brian-d1018/bis557 documentation built on Dec. 17, 2020, 6:21 p.m.