NightlordTW

################################################################################
# Calculates the maximum likelihood estimates of a logistic regression model
# Slopes are constrained to non-negative values
#
# fmla : model formula
# x : a [n x p] dataframe with the data. Factors should be coded accordingly
#
# OUTPUT
# beta : the estimated regression coefficients
# vcov : the variane-covariance matrix

NightlordTW

################################################################################
# Calculates the maximum likelihood estimates of a logistic regression model
#
# fmla : model formula
# x : a [n x p] dataframe with the data. Factors should be coded accordingly
#
# OUTPUT
# beta : the estimated regression coefficients
# vcov : the variane-covariance matrix
# ll : -2ln L (deviance)

NightlordTW

################################################################################
# Calculates the maximum likelihood estimates of a logistic regression model
#
# fmla : model formula
# x : a [n x p] dataframe with the data. Factors should be coded accordingly
#
# OUTPUT
# beta : the estimated regression coefficients
# vcov : the variane-covariance matrix
# ll : -2ln L (deviance)

NightlordTW

mylogit = glm(admit~gre+gpa+as.factor(rank), family=binomial, data=mydata)

NightlordTW

mydata = read.csv(url('http://www.ats.ucla.edu/stat/r/dae/binary.csv'))

NightlordTW

mydata$rank = factor(mydata$rank) #Treat rank as a categorical variable
fmla = as.formula("admit~gre+gpa+rank") #Create model formula
mylogit = mle.logreg(fmla, mydata) #Estimate coefficients
mylogit