Statistics

Using 2D Normal Mixtures for Spatial Point Patterns

This post describes the design and usage of the sppmix package, which defines classes and methods for spatial point pattern data and mixture models. You can install the package by this command. devtools::install_github("wangyuchen/sppmix") A two dimensional normal mixture is a basic building block of this package. We created a S3 class normmix and related methods for operations with normal mixture. In this post, you’ll learn everything that related to a normmix object.

Logistic Regression in R with Single-Trial Data

The examples in this post is taken from a class I took at the University of Missouri. These examples are originally provided in SAS and I translated them to R. The relationship between countries’ credit ratings and the volatility of the countries’ stock markets was examined in the Journal of Portfolio Management (Spring 1996). Our interest lies in whether volatility (standard deviation of stock returns) and credit rating (in percent) can be used to classify a country into one of the two market types (developed or emerging).

Nonlinear Regression in R

The examples in this post is taken from a class I took at the University of Missouri. These examples are originally provided in SAS and I translated them to R. Non-linear Regression with Available Starting Values The velocity of a chemical reaction (\(y\)) is modeled as a function of the concentration of the chemical (\(x\)). There are a total of 18 observations. The desired model is \[ y_i = \frac{\theta_0 x_i}{\theta_1 + x_i} + \epsilon_i.

Introduction to Meta-Analysis

This is a workshop on meta-analysis at East China Normal University. The slides of this talk is written in markdown and reveal.js. You can find the slides at https://www.prioritydetails.com/slides/meta-analysis/