Update finalproject

finalproject

@@ -24,7 +24,8 @@ data$X30.5drinks <- str_remove_all(data$X30.5drinks, " or more")
 data$X30.5drinks[data$X30.5drinks == "3 to 5"] <- 4
 data$X30.5drinks[data$X30.5drinks == "6 to 9"] <- 7.5
 data$X30.5drinks[data$X30.5drinks == "10 to 19"] <- 14.5
-mean(as.numeric(data$X30.5drinks))
+data$X30.5drinks <- as.numeric(data$X30.5drinks)
+mean(data$X30.5drinks)
 
 data$X30.cig <- str_remove_all(data$X30.cig, " days")
 data$X30.cig <- str_remove_all(data$X30.cig, " day")
@@ -34,7 +35,8 @@ data$X30.cig[data$X30.cig == "3-5"] <- 4
 data$X30.cig[data$X30.cig == "6-9"] <- 7.5
 data$X30.cig[data$X30.cig == "10-19"] <- 14.5
 data$X30.cig[data$X30.cig == "20-29"] <- 24.5
-mean(as.numeric(data$X30.cig))
+data$X30.cig <- as.numeric(data$X30.cig)
+mean(data$X30.cig)
 
 data$X30drink <- str_remove_all(data$X30drink, " days")
 data$X30drink <- str_remove_all(data$X30drink, " day")
@@ -45,7 +47,8 @@ data$X30drink[data$X30drink == "6-9"] <- 7.5
 data$X30drink[data$X30drink == "10-19"] <- 14.5
 data$X30drink[data$X30drink == "20-29"] <- 24.5
 data <- na.omit(data)
-mean(as.numeric(data$X30drink))
+data$X30drink <- as.numeric(data$X30drink)
+mean(data$X30drink)
 
 data$X30marijuana <- str_remove_all(data$X30marijuana, " days")
 data$X30marijuana <- str_remove_all(data$X30marijuana, " day")
@@ -56,8 +59,11 @@ data$X30marijuana[data$X30marijuana == "6-9"] <- 7.5
 data$X30marijuana[data$X30marijuana == "10-19"] <- 14.5
 data$X30marijuana[data$X30marijuana == "20-29"] <- 24.5
 data <- na.omit(data)
-mean(as.numeric(data$X30marijuana))
+data$X30marijuana <- as.numeric(data$X30marijuana)
+mean(data$X30marijuana)
+```
 
+```{r}
 # response data cleaning
 data$times.moved[data$times.moved == "None"] <- 0
 data$times.moved[data$times.moved == "Once"] <- 1
@@ -170,13 +176,22 @@ mean(as.numeric(data$N.safe))
 ```
 
 ```{r}
-# new response binary variable
+# response data exploration
+
+hist(as.numeric(data$X30.5drinks))
+hist(as.numeric(data$X30.cig))
+hist(as.numeric(data$X30drink))
+hist(as.numeric(data$X30marijuana))
+
+summary(data)
 
-data$substance <- 
 
 
 ```
 
+```{r}
+
+```