Skip to main content

Summer R Bootcamp

Problem Set 2

 

You will be working with the datasets containing demographic information of midwest counties from 2000 US census. Dataset_1, Dataset_2, and Dataset_3 are available on Courseworks.

 

Dataset_1 contains the following variables:

  • PID - Unique county identifier
  • county - County name
  • area - Area of county (units unknown). This variable is spread across 5 columns (IL - WI)
  • poptotal - Total population
  • state - State to which county belongs to

 

Dataset_2 contains the following variables:

  • PID - Unique county identifier
  • county - County name
  • race - Total population for each racial group (White, Black, American Indians, Asians, Other races)
  • inmetro - A factor with levels 0 and 1 to indicate whether county considered in a metro area
  • category - Miscellaneous

 

Dataset_3 contains the following variables:

  • PID - Unique county identifier
  • county - County name
  • popadults - Number of adults
  • perchsd - Percent of population with high school diploma
  • percollege - Percent of college educated population
  • percprof - Percent of population with professional degree
  • percbelowpoverty - Percent of population below poverty line
  • percchildbelowpovert - Percent of children below poverty line
  • percadultpoverty - Percent of adults below poverty line
  • percelderlypoverty - Percent of elderly below poverty line

 

Note: The original description of the dataset was not documented and the current descriptions here (provided by tidyverse package) are based on speculation.

 

Problem 1

  1. In Dataset_1 the area variable is spread across multiple columns that represent values of the state variable. Transform/Tidy the dataset as follows: move the column names (IL-WI) to the state variable and the column values to the area variable. Then, remove all the missing value from the dataset. Name the new dataset data1.

 

  1. In Dataset_2 the race variable combines information about the total population for each racial group, given in following order: White, Black, American Indians, Asians, Other. Pull apart this column into multiple columns, by splitting wherever a separator character appers. Name these variables popwhite, popblack, popamerindian, popasian, and popother, respectively. Name the new dataset data2.

 

  1. Merge data1 and data2: they should be joined by the state variable and the output result should contain all the observation from data1, regardless of whether they match or not. What are the dimensions of the output? How many missing values does it contain?

    Repeat the above procedure again by joining the datasets by the state and county variables.

 

  1. Merge data1 and data2: they should be joined by the state variable and the output result should contain all the observation from data2, regardless of whether they match or not. What are the dimensions of the output? How many missing values does it contain?

    Repeat the above procedure again by joining the datasets by the state and county variables.

 

  1. Merge data1 and data2: they should be joined by the state variable and the output result should contain all the observation from both data1 and data2, regardless of whether they match or not. What are the dimensions of the output? How many missing values does it contain?

    Repeat the above procedure again by joining the datasets by the state and county variables.

 

  1. Join data1 and data2 by the state and county variables such that the unmatched rows in either input dataset are not included in the result. Name the output dataset data.

 

  1. Remove PID and category from data by selecting all variables except these two. Then, remove all duplicates from the dataset.

 

  1. Add a new variable to the dataset (name it popdensity), which represents a population density in each county (in order to compute a population density you need to divide the total population by the corresponding area).

    Then, for each racial group, add a new variable to the dataset that will represent the percentage of the total population that belongs to this group (name them percwhite, percblack, percamerindan, percasian, and percother, respectively).

 

  1. Join data and Dataset_3 by the state and county variables such that the unmatched rows in either input dataset are not included in the result.

 

  1. Add new variables to the dataset that correspond to the percbelowpoverty and percadultpoverty variables and represent the actual number of people that belong to these groups. Name these variables popcbelowpoverty, and popcadultpoverty, respectively.

 

  1. Move the inmetro variable to the end of the dataset; change the position of the poptotal variable so that it comes after the area variables; change the position of the popdensity variable so that it comes after the poptotal variables.

 

  1. What is the biggest/smallest state by the total population? What is the biggest/smallest state by the total area? What is the average population density in each state? In each state, compare the population densities of counties that are in metro area to counties that are not. Hint: you might want to use the grouping functions.

 

  1. In all states combined, what is the average percentage of each racial group? In each state separately, what is the average percentage of each racial group? In each state separately, compare the average percentage of each racial group in metro areas to non-metro areas. Discuss the results you obtained.

 

  1. For each, state calculate the average percentage of the total population who are below poverty. Now using the same measurement, compare the metro areas to non-metro areas in each state. What do you observe? Now filter out counties with less than 50000 population and repeat the previous step. What do you observe?