Simple R script for performing logistic regression on binary categorical variables

logistic_regression

library(readr)
library(caret)

# Enter csv file path. If the data is in xls or xlsx, you will need to use read.table or the read_table function
# in the readr package
data <- read_csv("") 
attach(data)
summary(data)

#Enter column name with categorical variables you want to model and predict between the empty quotations.
names(data)[names(data) == ""] <- "category"

# Create factor for categorical variables.
data$category.f <- factor(data$category)

# Split data into training and test sets using 60% of the data.
inTrain <- createDataPartition(y = data$category.f, p = .60, list = FALSE)
training <- data[inTrain,]
testing <- data[-inTrain,]

# Fit logistic model to the data. Change 'Predictor' values to variable names in data
data.fit = glm(category.f ~ Predictor1 + Predictor2 + Predictor_n, data=training, family=binomial)
summary(data.fit)
data.prob = predict(data.fit, testing, type="response")

# Create prediction estimates to measure performance of model. Change "F" and "M" to category names.
data.pred = rep("F", dim(training)[1])
data.pred[data.prob > .5] = "M"
table(data.pred, training$category.f)
mean(data.pred == training$category.f)
mean(data.pred != training$category.f)

# Predict probability of simulated data falling under particular category. Change Variable1 and Variable2
# values to new data you want to predict. The values represent x,y pairs; for example {(2.8, 13), (1.8, 7)}
predict(data.fit, newdata=data.frame(Variable1=c(2.8, 1.8), Variable2=c(13, 7)), type="response")