Logistic Regression Assignment
There are 3 parts to this assignment that must be done in R-studio, please complete ALL parts.
This script will look a little bit different than the last one -- it is going to build upon your R skills that you've already developed.
I strongly recommend using the code we reviewed in class as an example for this -- use that to check your code and make sure you are on the right track.
a. Create a new dataset which has shorter name and same information.
b. Run frequencies and look at the following variables.
c. Show how the various variables differ by levels of the DFREE variable. We use CrossTable () to show the relationship between two categorical variables.
The IV Drug Use History at Admission variable has been created as a dummy variable as opposed to categorical. Look for ivhx2 and ivcx 3 in your dataset, you can use these dummy variables separately when you build your model.
Let us create a Naive model, meaning what are the probabilities of staying drug free without any predictors.
CONTINUE TO BUILD YOUR MODEL, ADDING THE VARIABLES YOU'VE DECIDED TO INCLUDE ONE BY ONE.
There are some interaction variables that are already written in the dataset. Try to apply those existed interaction variables or the interaction variable you created.
keep track of the odds radios, the significance of the variables you add, the model significance (modelsig), and the diagnostics (logisticPseudoR2s).
Attachment:- Assignment.rar