DESCRIPTION

Students:  Please note instructions in paper copy of quiz.  Save your
files often, make sure OMSI fills your entire screen at all times, etc.

A question may not fully fit into your OMSI question box, which is not
scrollable.  You can try adjusting the relative size of the question and
answer boxes, but remember that clicking CopyQtoA will copy the entire
question box to the answer box.

QUESTION

(25 pts, text) In the MovieLens genre data, each genre is coded as a
dummy variable, 1 if the movie belongs to that genre, 0 if not.  There
are 19 genres in all, including one named 'unknown'.  (The data treats
the latter as a real genre, so that 19 dummy variables are coded in each
row of u.item.) We do not exclude any dummies, and in calling lm(), for
instance, we would use all 19 dummies.  Why does not induce a matrix
rank problem?  (Assume no roundoff error.) 

QUESTION -ext .R -run 'Rscript omsi_answer2.R'

(25 pts, code) Within the 'lme4' package, which is included in 'rectools',
there is a data frame 'VerbAgg', one of whose components is 'item'.
Write code to print out how many different values 'item' can take on.

library(lme4)
print(        )

QUESTION -ext .R -run 'Rscript omsi_answer3.R'

(25 pts, code) Consider the InstEval data.  Say we are predicting rating
from 'studage' and and 'service', and we want to allow for the effect of
studage to differ according to whether the course is a service course.
Fill in the blanks.

library(rectools)
getInstEval()
ivl <- ivl[,       ]

print(lm(ivl$y ~ .,data=ivl1))

QUESTION -ext .R -run 'Rscript omsi_answer4.R'

(25 pts, code) Continue to assume the setting of Problem 3, but 
now fitting a fully quadratic model.  Fill in the blanks.  Do NOT use
'polyreg'.

library(rectools)
getInstEval()
ivl1 <- ivl[,    ]

print(lm(ivl$y ~ .,data=ivl1))