Upload New File

SS19-Erklärung_unterschiedlicher_Gesundheitspolitiken_in_Europa/_Example__Linear_regression_models.r

+#---------------#
+#### Options ####
+#---------------#
+
+  #Clear workingspace
+    rm(list = ls())
+  
+  #load packages
+    if (!require("csv")) install.packages("csv") # usage of read.csv2()
+    library(csv)
+
+    if (!require("rstudioapi")) install.packages("rstudioapi") #usage of getActiveDocumentContext()
+    library(rstudioapi)
+
+  
+  #Set Working Directory
+    setwd(dirname(rstudioapi::getActiveDocumentContext()$path)) # setting working directory to source file location
+
+
+
+
+#------------------------#
+#### Loading datasets ####
+#------------------------#
+  
+  data <- read.csv2("OECD.ComparativeWelfare.data/OECD.ComparativeWelfare.data.csv")
+  
+  
+#-----------------------------------------#
+#### Overview / Descriptive Statistics ####
+#-----------------------------------------#  
+  
+  #Overview (first six observations of the dataset)
+    head(data)
+  
+  #Descriptive Statistics for all variables
+    if(!require("Hmisc")) install.packages("Hmisc")
+    library(Hmisc)
+    describe(data) 
+  
+  #Descriptive Statistics for all variables
+    if(!require("psych")) install.packages("psych")
+    library(psych)
+    describe(data)
+    describeBy(data, data$Country)
+  
+  #Mean of variable  
+    mean(data$leftmaj) # mean is "NA" because there are missing values on this variable
+    mean(data$leftmaj, na.rm = TRUE) # "na.rm = TRUE" removes the missing values in the calculation of the mean value
+  
+  #Summary of the results of various model fitting functions  
+    summary(data$leftmaj, na.rm = TRUE)
+    
+  #Summary of partial data set  (Only statistics for "Germany")
+    summary(data$leftmaj[data$Country=="Germany"], na.rm = TRUE)
+  
+  #Graphical representation of the relationship between two variables
+    plot(data$leftmaj , data$Drug.use.disorders)
+    
+    
+#-------------------------------#
+#### Linear Regression Model ####
+#-------------------------------#   
+    
+    model1 <- lm(data$Share.GDP.gov.expenditures.health ~ data$Years.lost) # error because the data type of the variable is not "numeric"
+    summary(model1) # doesn't work because of error above (-> Data type of one or more variables is not numeric)
+    
+    #Check in type of data (TRUE: numeric / FALSE: not numeric)
+      for(i in 3:length(names(data))-2) {
+        print(names(data[i]))
+        print(is.numeric( eval(parse(text = paste("data$",names(data[i]), sep="" )))))
+        print("------------")
+      }
+    
+    #saving the needed variables as numeric
+      data$Years.lost <- as.numeric(data$Years.lost)
+      data$Share.GDP.gov.expenditures.health <- as.numeric(data$Share.GDP.gov.expenditures.health)
+    
+    #Now is the regression working
+      model1 <- lm(data$Share.GDP.gov.expenditures.health ~ data$Years.lost)
+      summary(model1)
+    
\ No newline at end of file