Neighborhood patterns

As our next task, we seek to find patterns between pickup and drop-off neighborhoods and other variables such as fare amount, trip distance, traffic and tipping. Since ultimately RevoScaleR is just another R package, it's important to know that to really put it to use, it must integrate and interact with other R packages to make our analysis possible. We've already seen examples of how to use packages such as lubridate or rgeos for data transformations with RevoScaleR, and how to use results returned by rxSummary or rxCrossTabs and pass them to other R functions. Sometimes, in the process of examining our results, we notice certain attributes about the data that need to be re-examined.

Learning objectives

At the end of this chapter, we will know how to

• let data summaries guide your decision about how the data should be formatted
• know when to use rxCrossTabs and when to use rxCube
• recognize which transformations should best be performed on the fly for the sake a this or that summary, and which one should be built into the data
• use the factor function or the alternative rxFactor function for "releveling" a factor column