R
How many stars represents a good P-Value
3 (***)
What does the update() function do? (what does it update?)
update a model (I.E. linear)
a first order polynomial fit of the data is what shape?
if the Regression Coefficient is 0.2 what does that tell us about the variables?
not correlated / moderatly weak
(T/F) As sample size increases more variation can be modeled
True
function for plotting a regressed line in R
abline()
How do you access data from a csv file
read.csv()
first you must download the csv into the project
are higher order polynomials always better
No
How do you transform a plot whos best fit line is exponential
Log the data
How many variables can predict question scores on stack overflow
(6, 7, 16, 12, 5)
7
code for printing out a pdf of a plot
dev.print(device=pdf, "name")
Code to remove null data values
na.omit(df)
if the Regression coeficcient is -0.9 what does that tell us about the data?
They have a strong negative correlation
a great P-value is less than what number?
0.001
what does a clustering model do?
clusters data into larger groups
Code to subdivide the plotting area into four regions
(hint: used for diagnostic test)
par(mfrow=c(2,2))
What is wrong with this code?
lm.fit=lm(medv, lstat)
lm.fit=lm(medv~lstat)
why do we use regression on data sets?
to predict data, and identify correlations
y=0.28x + 9.916
x=0.3
what is y?
9.95 or 10.00 or 10.05 or 10.10
10
The Full meaning of CDR's
hint: Used in the Sri Lanka Paper
Call Detail Records
how do you get confidence intervals
predict(data.frame(c(30)), interval="confidence")
What is wrong with this code?
abline(lm.fit,lwd=3)
abline(lm.fit,lwd=3,col="red")
plot(lstat,medv,col="red")
plot(lstat,medv,pch=20)
plot(lstat,medv,pch='+")
plot(1:20,1:20,pch=1:20)
plot(lstat,medv,pch="+")
A term used to describe the case when the predictors in a multiple regression model are correlated
Multicollinearity
Description of Heteroscedasticity
Linear Regression with varying error terms
(T/F) a low R value means that there is no statistical relationship between variables
False