Week 6

This week we learned more about logistic regression, which is used when you have a binomial response variable, like yes or no. Pretty much the same as last week, I’ll put in a few examples below that we did.

Here is a problem we did for class on beer.

beer <- read.csv("http://www.cknudson.com/data/MNbeer.csv")
head(beer)

##       Brewery         Beer              Description Style ABV IBU Rating Good
## 1     Bauhaus  Wonderstuff     New Bohemian Pilsner Lager 5.4  48     88    0
## 2     Bauhaus    Stargazer German Style Schwarzbier Lager 5.0  28     87    0
## 3     Bauhaus  Wagon Party    West Cost Style Lager Lager 5.4  55     86    0
## 4     Bauhaus    Sky-Five!        Midwest Coast IPA   IPA 6.7  70     86    0
## 5 Bent Paddle         Kanu         Session Pale Ale   Ale 4.8  48     85    0
## 6 Bent Paddle Venture Pils            Pilsner Lager Lager 5.0  38     87    0

library(faraway)

Holding IBU constant, beers with higher ABVs are more likely to be good with a multiplicative change of 1.008282 in the logodds of being a good beer. After accounting for IBU, there is a relationships between the ABV and a beer’s log odds of being “Good” at the significance level .1. The pvalue for ABV is .0697.

modBeer <- glm(Good ~ ABV + IBU, family = binomial, data = beer)    
exp(coef(modBeer))

##  (Intercept)          ABV          IBU 
## 9.072303e-05 3.702429e+00 1.008282e+00

summary(modBeer)

## 
## Call:
## glm(formula = Good ~ ABV + IBU, family = binomial, data = beer)
## 
## Deviance Residuals: 
##     Min       1Q   Median       3Q      Max  
## -1.2930  -0.7759  -0.4435   0.7678   2.1315  
## 
## Coefficients:
##              Estimate Std. Error z value Pr(>|z|)  
## (Intercept) -9.307699   3.678245  -2.530   0.0114 *
## ABV          1.308989   0.721717   1.814   0.0697 .
## IBU          0.008248   0.025139   0.328   0.7428  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for binomial family taken to be 1)
## 
##     Null deviance: 51.564  on 43  degrees of freedom
## Residual deviance: 42.000  on 41  degrees of freedom
## AIC: 48
## 
## Number of Fisher Scoring iterations: 5

ilogit(-9.307699+1.308989*4.2 + .008248*27)

## [1] 0.02692911

For more explanation on this project, we made a powerpoint which can be accessed here: https://docs.google.com/presentation/d/1cAaw2Xo1nnEJheXWdmkaHpF99wlxy0Knvc9C4KNfzlw/edit?usp=sharing

That’s pretty much all for this week, we mostly worked on homework