clean up RailTrails data set

[rpruim]

Things to do:

• add season variable with three levels
• make factor version of weekday variable (perhaps day with levels weekday and weekend)
• leave current 0/1 indicators there for backward compatibility

[rpruim]

Perhaps also add a richer version (RailTrail2) with finer data resolution (date, holiday or not, etc.)

[nicholasjhorton]

That's a good idea: I'll try to make it happen (but it will need to wait for a release in November or later).

[rpruim]

Danny and I created a data frame called Riders. It's in the github repo. Take a look at it to see whether you have suggestions for improvement.

head(Riders)
##         date       day highT lowT hi lo precip clouds riders  ct weekday wday
## 1 2005-04-05   Tuesday    62   39 62 39      0    0.8    236 236       Y    Y
## 2 2005-04-06 Wednesday    75   43 75 43      0    6.3    156 156       Y    Y
## 3 2005-04-07  Thursday    70   47 70 47      0    9.9    328 328       Y    Y
## 4 2005-04-08    Friday    65   48 65 48      0    7.2    418 418       Y    Y
## 5 2005-04-09  Saturday    66   38 66 38      0    0.0    629 629       N    N
## 6 2005-04-10    Sunday    74   33 74 33      0    2.5    635 635       N    N

The redundant variables are for the sake of the mosaic cheat sheet (to keep the code shorter).

[rpruim]

Unless there is some other data to add to this, I think it is ready to go.