lr: Linear Regression Fitting
In XuelinGu/LinearRegression: Linear Regression Fitting

Description Usage Arguments Details Value Examples

View source: R/LinearRegressionFitting.R

lr is used for fitting linear regression models with numeric or categorical covariates. It can be used to estimate regression coefficients with least square method and make inference based on relevant t-test and F-test (sequential F-test can be realized by F_test).

1	lr(formula, data, coding = "reference", intercept = TRUE, reference = 1)

`formula`	an object of class `formula`: a symbolic form of the model to be fitted which should contain both respond and covariate variables. The details of model specification form are given under "Details".
`data`	an optional data frame,list or environment containing the variables in the model. If not found in data or by default, the variables would be taken from environment where ls is called.
`coding`	optional,the method to be used for fitting categorical covaraites in the model. There are two options: coding = "reference" (by default) or "means" (see "Details").
`intercept`	logical. If TRUE (by default), the corresponding fitting model will contain intercept term.
`reference`	an optional categorical covariate group which would be considered as reference group when `coding` = "reference" (by default). If not specified, model fitting will take first group occured among the corresponding covariate as reference.

Models for lr are specified symbolically like "y ~ x1 + x2". A typical model has the form "response ~ covariates" where response is the numeric response vector and covariates are a series of numeric or categorical terms which specifies a linear predictor for response. A covariate specification of the form "x1 + x2" indicates covariates will contain all the observations in "x1" and "x2".

In addition, regarding to the variable in a data frame contianed in model formula, they can be either expressed as mtcars$mpg or mpg with data = mtcars where "mtcars" is the name of a data frame, and "mpg" is the variable name in "mtcars".

coding method is used to cope with model containing categorical covariates. Two common methods: "cell reference coding" and "cell means coding" are supported by setting coding = "reference" (by default) or "means". Former one takes one group of the corresponding categorical covariate as a reference group and reserve intercept in the model, while latter one just eliminates intercept in the model.

linear regression fitting coefficients and reference results.

## A example for numeric respond and covariate variables
y = c(23, 24, 26, 37, 38, 25, 36, 40)
x1 = c(1, 2, 3, 4, 5, 6, 7, 8)
x2 = c(23, 32, 34, 20, 24, 56, 34, 24)
result = lr(y ~ x1 + x2)

## A example for numeric respond and categorical variables
x3 = c("M", "F", "F", "U", "M", "F", "F", "U")
result = lr(y ~ x1 + x2 + x3)

## A example for variables in data frame
result = lr(mpg ~ cyl + disp + hp, data = mtcars)
result = lr(mtcars$mpg ~ mtcars$cyl + mtcars$disp + mtcars$hp)