bike-share-prediction

Purpose

The purpose of this project is to demonstrate how to build and use predictive models and automate Markdown reporting using R. We use the bike share dataset and automate the exploratory analysis and predictive modeling for the number of bike rentals per day for each day of the week.

Required Packages

The following packages are required for this project:

• readr: reading in CSV files containing data
• tidyverse: general data manipulation and plotting
• cowplot: plotting in grids
• caret: testing and making machine learning models
• rmarkdown: rendering Rmd file output files

Output Files

• The analysis for Monday is available here
• The analysis for Tueday is available here
• The analysis for Wednesday is available here
• The analysis for Thursday is available here
• The analysis for Friday is available here
• The analysis for Saturday is available here
• The analysis for Sunday is available here

Automation

We automate the analysis process in the render.R file, which contains the code below. It loops through the weekday ID numbers and runs the analysis and puts the rendered file in the Reports folder.

# Create a list of weekday names in the same order as the weekday ids. 
weekdayList <- list("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday",
                    "Friday", "Saturday")

# Loop through the weekday indexes going from Sunday to Saturday.
for(weekday in c(0,1,2,3,4,5,6)){
  # Render Analysis.Rmd file...
  rmarkdown::render(
    "Analysis.Rmd",
    # and put the output in the Reports folder naming the file as the 
    # corresponding weekday by adding 1 to the id to get the weekday name for
    # the file name...
    output_file = paste0("Reports/", weekdayList[[weekday+1]]),
    # and pass the weekday id as the "weekday" parameter and the name of the 
    # weekday as the "dayName" parameter..
    params=list(
      weekday=weekday,
      dayName=weekdayList[[weekday+1]]
      )
    )
}