Write an R function prism(df, decisionAttributeIndex) that applies the Prism alg

Question

Write an R function prism(df, decisionAttributeIndex) that applies the Prism algorithm to the dataset specified by the dataframe df where decisionAttributeIndex is the (1-based) index that corresponds to the decision attribute in the dataframe df. The function should output the rules that are found. Show that your function works by running it on the two datasets given below:

Decision attribute is f (i.e., attribute number 5)
Output should be:
If b = L then f = 0
If c = 1 then f = 1 (instead of c = 1, could be a = 0)
If a = 1 then f = 2 (instead of a = 1, could be b = R, c = 0, or d = L)
If a = 2 then f = 3 (instead of a = 2, could be b = S, c = 2, or d = H)
Note: Your output doesn’t have to list all the alternatives for conditions in a rule; we’re just showing them so you can check for alternate answers

Decision attribute is isMammal (i.e., attribute number 5)
Output should be:
if fourLegged = no then isMammal = no
if bodyTemp = warmBlooded then isMammal = yes (instead of bodyTemp = warmBlooded, could be laysEggs = no, fourLegged = yes, or hibernates = no)
This dataset is available on Canvas under Files/Lecture Notes/Lecture Notes for Week 4 as dataset2.csv
You should write your function using a text editor like Notepad or Wordpad, and save it as a text file. The very first line in your file should be a comment that identifies the authors (no more than 2) of the program.
If your file is named “prism.r”, we should be able to do the following to run it from within the R interpreter:
source(“prism.r”)
df = read.csv("c:\dataset1.csv", header = TRUE, sep=",")
prism(df, 5);

Explanation / Answer

No need to create an object , you can just create a list like this:

Here an example using cclust(you should use your matlab cluster here) to show that the 4 variables are enough to use clustIndex function:

Related Questions

Navigate

• Browse (All)
• Browse W
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.