Log day 17

Concepts

In class we learned how to make increasing seasonal variation constant using transformations. We also learned how to model this data using dummy variables.

Increasing or Constant Seasonal Variation

For this example I will use the data set airpass from the R package faraway. Airpass contains the number of passengers (in thousands) traveling by plane per month from 1949 to 1951.

library(faraway)

## Warning: package 'faraway' was built under R version 3.4.4

data(airpass)
plot(pass~year, data = airpass, type = "l")

This data has increasing seasonal variation. Now we transform this data to have constant variation.

plot(sqrt(pass)~year, data = airpass, type = "l")

plot(log(pass)~year, data = airpass, type = "l")

As you can see the log transformation makes the data look the closest to constant seasonal variation.

Dummy variables

Now we can use dummy variables to group our year factor by months.

head(airpass)

##   pass     year
## 1  112 49.08333
## 2  118 49.16667
## 3  132 49.25000
## 4  129 49.33333
## 5  121 49.41667
## 6  135 49.50000

justyear=floor(airpass$year)
modecimal=airpass$year - justyear
mofactor=factor(round(modecimal*12))
head(cbind(airpass$year, mofactor))

##               mofactor
## [1,] 49.08333        2
## [2,] 49.16667        3
## [3,] 49.25000        4
## [4,] 49.33333        5
## [5,] 49.41667        6
## [6,] 49.50000        7

levels(mofactor)=c("Jan", "Feb", "Mar", "Apr", "May", 
                      "Jun", "Jul", "Aug", "Sep", "Oct",
                      "Nov", "Dec")
airpass$justyear=justyear
airpass$mofacto=mofactor

Now that the data is grouped by months we can create a model.

mmod=lm(log(pass)~justyear + mofactor, data = airpass)

Finally we can plot our model against our original.

plot(log(pass)~year, data = airpass, type = "l")
lines(airpass$year, mmod$fitted.values, type = "l", col = "blue")

As you can see the dummy variable data set fits our data fairly well.

Peer Review

I also received some feeback in class for our group project. The biggest concern my partner had was that our paper read more as an R guide then as a paper. We will have to add our r code into a appendix and not in the paper itself. We also need to elabirate on why our data maters and connect it back to our research topic more. I think the most important thing that I learned from this peer review is how to present our data in a paper form instead of a how to R guide.