search

ddsjoberg / sjosmooth

Kernel-smoothed Regression Estimation
http://www.danieldsjoberg.com/sjosmooth/
Other
2 stars 0 forks source link

readme

Travis build status AppVeyor build status Coverage status CRAN status

sjosmooth

The goal of sjosmooth is to provide kernel smoothed estimates for time to event data.

Installation

You can install the released version of sjosmooth from GitHub with:

remotes::install_github("ddsjoberg/sjosmooth")

Functions

The sjosmooth package has two primary functions. The first calculates kernel-weighted regression models (sm_regression). The user specifies a regression model and a variable for weighting, and sm_regression will estimate a weighted regression model for each unique value of the specified variable.

The second function (sm_predict) calculates kernel-weighted predictions from regression models (i.e. outcomes that can be calculated from the predict function).

The package was written generally, such that any regression model should be compatible. However, only survival::coxph(), and survival::survreg(), stats::lm(), and stats::glm() models were tested.