bis557
bis557 - A Package for Computational Statistics
This package contains course materials from Yale BIS557 Computational Statistics (2020 Fall). The goal is to design and implement algorithms for statistical analyses, including regression models (ridge/lasso/multi-logistic), cross-validation, stochastic gradient descent, neural networks, etc. R and Python code are designed for the implementation.
Installation
You can install the development version from GitHub with:
# install.packages("devtools")
devtools::install_github("yijunyang/bis557")
Example
This is a basic example which shows you how to solve a numerically-stable ridge regression that takes into account colinear regression variables:
# library(bis557)
data("iris")
# create a colinear term
iris$colinear <- 2*iris$Petal.Width + 0.1
# claim formula and dataset
form <- Sepal.Length ~ .
d <- iris
# make model matrices
mms <- make_model_matrices(form, d, contrasts = NULL)
X <- mms$X
Y <- mms$Y
# implement ridge regression
ridge_regression(form, d, lambda = 10)
# run the ridge regression function with cross validation
cv_ridge(form, d, lambda = seq(0, 0.05 ,0.001))