bglm: Function to carry out generalized linear regression on a...

In bigReg: Generalized Linear Models (GLM) for Large Data Sets

Description

Function to carry out generalized linear regression on a data_frame data object

Usage

bglm(formula, family = gaussian_(), data, weights = NULL, offset = NULL,
  start = NULL, control = list(), etastart = NULL, mustart = NULL)

Arguments

formula	formula that defines your regression model
family	family object from activeReg, e.g. .gaussian(), .binomial(), .poisson(), .quasipoisson(), .quasibinomial(), .Gamma(), .inverse.gaussian(), .quasi()
data	data_frame object containing data for linear regression
weights	weights for the model
offset	offsets for the model
start	starting values for the linear predictor
control	list of parameters for .control() function
etastart	starting values for the linear predictor
mustart	starting values for vector of means

Examples

require(parallel)
data("plasma", package = "bigReg")
plasma1 <- plasma
plasma1 <- data_frame(plasma1, 10, path = "outputs", nCores = 1)
plasma_glm <- bglm(ESR ~ fibrinogen + globulin, data = plasma1, family = binomial_("logit"))
summary(plasma_glm)

Example output

Loading required package: Rcpp
Loading required package: parallel
Loading required package: uuid
Loading required package: MASS

Call:
 bglm(formula = ESR ~ fibrinogen + globulin, family = binomial_("logit"), 
    data = plasma1) 
            Estimate Std.Error  z.value P(>|z|)
(Intercept) -12.7921    5.7964  -2.2069   0.027
fibrinogen    1.9104    0.9710   1.9674   0.049
globulin      0.1558    0.1195   1.3032   0.193

logLik: -42.83287 , df: 4
AIC: 91.66573 , BIC: 106.4602 

bigReg documentation built on May 2, 2019, 6:43 a.m.