#Dependent variable:
#MAP: 1 indicates abnormal and 0 indicates
normal
#Independent variable:
#Bloodweight, continuous variable
#White blood cell in thousands, continuous
variable
#MCV, continuous variable
#Platelets, continuous variable
#Blood Urea Nitrogen, continuous variable
#Glucose, continuous variable
#Creatinine, continuous variable
#Total cholesterol, continuous variable
#Triglycerides, continuous variable
#Hdl cholesterol, continuous variable
#Ldl cholesterol, continuous variable
#C-Reactive protein, continuous variable
#Glycated Hemoglobin, continuous variable
#Uric acid, continuous variable
#Hematocrit, continuous variable
#Hemoglobin, continuous variable
#Cystatin C, continuous variable
#The dependent variable was made binary
data$MAP<-factor(data$MAP,
levels = c(0,1),
labels
=
c("anomalous","unanomalous"))
library(readr)
data <-read_csv("MAP.csv")
#Determine missing values
is.na(data)
#Delete the missing values
data <-na.omit(data)
#View data
View(data)
#View variable names
names(data)
#View the basic statistics of each variable in
dataset 'data
summary(data)
#View variable types
str(data)
#Put all independent variables into the
model
model.MAP
glm(MAP~.,data=data,family=binomial())
#View the result of model
summary(model.MAP)$coefficients
#Variables have statistically significance
#Calculate odds ratios and their confidence
intervals
exp(cbind("OR"=coef(model.MAP),confint(
model.MAP)))
test_data <- read_csv("test_data.csv")
test_data$MAP <- ifelse(test_data$MAP ==
"unanomalous", 0, 1)
test_data_clean <-na.omit(test_data)
predictions <- predict(model.MAP, newdata
= test_data_clean, type = "response")
table(Predicted = ifelse(predictions > 0.5, 1,
0), Actual = test_data_clean$MAP)
