IE6600 - hw2
Due on 2/21/2022 11:59pm PT
Zhenyuan Lu
IE6600 Homework Instructions
You should process your homework in R Markdown (.Rmd), and knit it to .pdf file (do not include any data
or other materials). Attentively check all the references I mentioned in the class or from other resources.
Once the homework is completed, you need to have your homework compressed into one .zip file
(hw2YourFullName.zip), and submit it to the assignment section on Canvas. In the .zip file, it should
contain the following documents:
hw2YourFullName.rmd
hw2YourFullName.pdf
Please include all your codes and results for each of the problem, and keep them organized and clear. All of
your codes should run successfully. Problems related to any plots/charts should be generated by ggplot2
mainly. If it’s neccessary, please deal with the missing values, overplotting, or labels on axis/legend properly.
All the solutions may vary.
1
Section A
Only use ggplot2 for plotting This section is for testing your ggplot2, and data exploration skills. Dataset
msleep from ggplot2 package will be using through this section. Use ? to check the documentation of msleep.
Problem 1
We are interested in those animals whose awake time over 12 hours. Create a bar chart as the following
figure. Remove the NA values from feeding types: carnivore, omnivore, insectivore and herbivore.
hints: You may adjust the angel of x-axis label by using theme(axis.text.x=element_text()), and the legend
labels by using scale_fill_discrete().
0
3
6
9
Arti
oda
ctyl
a
Car
nivo
ra
Cet
ace
a
Dip
roto
don
tia
Erin
ace
omo
rpha
Hyr
aco
idea
Lag
omo
rpha
Mon
otre
mat
a
Per
isso
dac
tyla
Prim
ates
Pro
bos
cide
a
Rod
enti
a
Sca
nde
ntia
Sor
icom
orph
a
Order
Co
un
t
Feeding Type
Carnivore
Herbivore
Insectivore
Omnivore
Problem 2
We would like to investigate how’s the relationship between total amount of sleep (hr) and brain weight(kg)
among feeding types: carnivore, omnivore, insectivore and herbivore. Plot total amount of sleep (hr)
versus brain weight (kg), applying color mapping on the feeding types(vore). Remove the NA group
from feeding types. Include a smoothing line on the plot. What do you notice in the plot?
Please deal with labels on axis/legend properly.
2
Problem 3
Still working on the above plot. Apply log transformation on the brain weight Brain Weight (Kg), Log,
what do you observe in the plot?
Section B
Only use ggplot2 for plotting
Section B uses FY 2019 H-1B Employer Data from U.S. Citizenship and Immigration Services. Download
FY2019 H-1B data from: https://www.uscis.gov/tools/reports-studies/h-1b-employer-data-hub-files
To read the data manual: https://www.uscis.gov/tools/reports-studies/understanding-our-h-1b-employer-
data-hub
The H-1B is a visa in the United States under the Immigration and Nationality Act, section 101(a)(15)(H)
that allows U.S. employers to temporarily employ foreign workers in specialty occupations. A specialty
occupation requires the application of specialized knowledge and a bachelor’s degree or the equivalent of
work experience.
Use read.csv() to import the dataset to R.
Problem 1
Import the H-1B data.
• You may notice the data types of “Initial.Approvals”, “Initial.Denials”, “Continuing.Approvals”, and
“Continuing.Denials” are wrong. We need to convert them into numerical columns.
• Return a data frame containing the top 5 employers which have the most cases of initial approved
H-1B. This data frame should have the columns: employer, initial approvals, initial denials, continuing
approvals, and continuing denials. Show the top 5 data frame.
• Plot a bar chart of Employer versus Initial approvals, mapping Initial Denials as fill, what do you
notice based on the plot?
Hint1: All the variables should be be associated with the proper data types
Hint2: Using function gsub() to eliminating the “,” for every three decimal places. e.g. 1,000 to 1000
Hint3: When converting data from factor to numeric, be aware of the values
Problem 2
Download geocode data: https://northeastern.instructure.com/courses/91043/files/11702036/download?
download_frd=1
If this link doesn’t work, please go to Canvas - Home - Homework - usZipGeo.csv
• Join H-1B data table with geocode data table by State and Zip columns.
• This new data frame should include columns: zip, employer, initial approvals, initial denials, continuing
approvals, continuing denials, state, city, longitude, and latitude.
• Insert a new column prop into this new data frame by the formula: inital denial/initial approval
Hint1: When joining two tables, make sure all the key column names are the same from both tables.
3
Problem 3
We are interested in the H-1B cases around Bay Area, California.
Create a map of the California, and then adjust the plotting x/y limits to a proper zoom level of Bay Area.
Then showing the locations of each employer along with, the prop less than 0.1 (mapped as the color/fill),
and the initial approvals (mapped as the size).
hints: Install maps and mapproj packages, and use the ggplot2::map_data() to draw “California” region
of the US. Using coord_map() to set up the view range of your map.
Answer may vary
The following example is just for your reference. If you plot is slightly different from it, you should be fine.
Bay Area
37.0
37.5
38.0
38.5
−122.75−122.50−122.25−122.00−121.75−121.50
Longitude
La
tit
ud
e
Initial Denials: Initial Approvals <0.1
TRUE
FALSE
Initial Approvals
0
1000
2000
3000
4