Part 1: Overview of last Classes
Part 2: More Logic and Condensing Data
- Part 2 Exercises
Part 3: Customizing Plots
- Part 3 Exercises
Part 4 Installing and loading R packages
- Part 4 Exercises

Part 1: Overview of last Classes

Reading in Data

Download new data set, frequency of several diseases in Boston since the 1880’s BosEpi.tab.txt

bosEpi = read.delim(file.choose())

View data table

View(bosEpi)

Use the str() and summary() to get a quick summary of the data

str(bosEpi)

'data.frame':   30062 obs. of  8 variables:
 $ disease: Factor w/ 23 levels "BRUCELLOSIS [UNDULANT FEVER]",..: 20 5 22 20 5 22 20 5 20 16 ...
 $ event  : Factor w/ 2 levels "CASES","DEATHS": 2 2 2 2 2 2 2 2 2 2 ...
 $ number : int  4 7 2 2 5 2 5 8 2 1 ...
 $ loc    : Factor w/ 1 level "BOSTON": 1 1 1 1 1 1 1 1 1 1 ...
 $ state  : Factor w/ 2 levels "IN","MA": 2 2 2 2 2 2 2 2 2 2 ...
 $ Year   : int  1888 1888 1888 1888 1888 1888 1888 1888 1888 1888 ...
 $ Month  : int  7 7 7 7 7 7 8 8 8 8 ...
 $ Day    : int  22 22 22 29 29 29 5 5 12 12 ...

summary(bosEpi)

                               disease         event      
 DIPHTHERIA                        : 3866   CASES :14620  
 TUBERCULOSIS [PHTHISIS PULMONALIS]: 3479   DEATHS:15442  
 TYPHOID FEVER [ENTERIC FEVER]     : 3312                 
 SCARLET FEVER                     : 3282                 
 MEASLES                           : 3164                 
 PNEUMONIA AND INFLUENZA           : 2276                 
 (Other)                           :10683                 
     number            loc        state           Year     
 Min.   :   0.00   BOSTON:30062   IN:    1   Min.   :1888  
 1st Qu.:   1.00                  MA:30061   1st Qu.:1913  
 Median :   7.00                             Median :1925  
 Mean   :  21.68                             Mean   :1927  
 3rd Qu.:  25.00                             3rd Qu.:1936  
 Max.   :2859.00                             Max.   :2013  
                                                           
     Month             Day       
 Min.   : 1.000   Min.   : 1.00  
 1st Qu.: 3.000   1st Qu.: 8.00  
 Median : 6.000   Median :16.00  
 Mean   : 6.462   Mean   :15.75  
 3rd Qu.: 9.000   3rd Qu.:23.00  
 Max.   :12.000   Max.   :31.00

Important columns of the data set
number - Number of occurences
loc - Location of disease
event - Whether the number in number is number of cases or number of deaths from disease
disease - Disease Name
Year - Year of occurence
Month - Month of occurrence
Day - Day of occurrence

Accesing Data

Can get a column by using $ and the name of the column or using [,] notation
Also since this is a very large table you can use head() to print a manageable amount of information

head(bosEpi$Year)

[1] 1888 1888 1888 1888 1888 1888

head(bosEpi[,6])

[1] 1888 1888 1888 1888 1888 1888

Get multiple columns using [] and :
The following will get the columns 6 through 8

head(bosEpi[,6:8])

  Year Month Day
1 1888     7  22
2 1888     7  22
3 1888     7  22
4 1888     7  29
5 1888     7  29
6 1888     7  29

Or get columns by name and []

head(bosEpi[,c("disease", "number", "event", "Year")])

                        disease number  event Year
1 TYPHOID FEVER [ENTERIC FEVER]      4 DEATHS 1888
2                    DIPHTHERIA      7 DEATHS 1888
3    WHOOPING COUGH [PERTUSSIS]      2 DEATHS 1888
4 TYPHOID FEVER [ENTERIC FEVER]      2 DEATHS 1888
5                    DIPHTHERIA      5 DEATHS 1888
6    WHOOPING COUGH [PERTUSSIS]      2 DEATHS 1888

Getting only select rows
For example to get first five rows of bosEpi

bosEpi[1:5,]

                        disease  event number    loc state
1 TYPHOID FEVER [ENTERIC FEVER] DEATHS      4 BOSTON    MA
2                    DIPHTHERIA DEATHS      7 BOSTON    MA
3    WHOOPING COUGH [PERTUSSIS] DEATHS      2 BOSTON    MA
4 TYPHOID FEVER [ENTERIC FEVER] DEATHS      2 BOSTON    MA
5                    DIPHTHERIA DEATHS      5 BOSTON    MA
  Year Month Day
1 1888     7  22
2 1888     7  22
3 1888     7  22
4 1888     7  29
5 1888     7  29

Combining getting only select rows and columns

Using column numbers

bosEpi[1:5,1:3]

                        disease  event number
1 TYPHOID FEVER [ENTERIC FEVER] DEATHS      4
2                    DIPHTHERIA DEATHS      7
3    WHOOPING COUGH [PERTUSSIS] DEATHS      2
4 TYPHOID FEVER [ENTERIC FEVER] DEATHS      2
5                    DIPHTHERIA DEATHS      5

Using column names

bosEpi[1:5,c("disease", "event", "number")]

                        disease  event number
1 TYPHOID FEVER [ENTERIC FEVER] DEATHS      4
2                    DIPHTHERIA DEATHS      7
3    WHOOPING COUGH [PERTUSSIS] DEATHS      2
4 TYPHOID FEVER [ENTERIC FEVER] DEATHS      2
5                    DIPHTHERIA DEATHS      5

Using Logic

Here is a table of logic operators to use on data

operator	meaning
<	less than
<=	less than or equal to
>	greater than
>=	greater than or equal to
==	exactly equal to
!=	not equal to

Testing the year column for the year 1888 and looking at the first 50 rows

head(bosEpi$Year == 1888, n = 50)

 [1]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[10]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[19]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[28]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[37]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[46] FALSE FALSE FALSE FALSE FALSE

Testing for anything happening before 1890 not including 1890

head(bosEpi$Year < 1890, n = 100)

  [1]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [10]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [19]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [28]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [37]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [46]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [55]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [64]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [73]  TRUE  TRUE  TRUE FALSE FALSE FALSE FALSE FALSE FALSE
 [82] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
 [91] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[100] FALSE

Testing for anything happening before 1890 including 1890

head(bosEpi$Year <= 1890, n = 300)

  [1]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [10]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [19]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [28]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [37]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [46]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [55]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [64]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [73]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [82]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [91]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[100]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[109]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[118]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[127]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[136]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[145]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[154]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[163]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[172]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[181]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[190]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[199]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[208]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[217]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[226]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[235]  TRUE  TRUE  TRUE FALSE FALSE FALSE FALSE FALSE FALSE
[244] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[253] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[262] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[271] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[280] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[289] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
[298] FALSE FALSE FALSE

Combining logic with accessing elements of a data.frame
Getting the rows that contain data from only the year 1888 by telling R to get rows only where bosEpi$Year is equal to 1888

bosEpi[bosEpi$Year == 1888,]

                         disease  event number    loc state
1  TYPHOID FEVER [ENTERIC FEVER] DEATHS      4 BOSTON    MA
2                     DIPHTHERIA DEATHS      7 BOSTON    MA
3     WHOOPING COUGH [PERTUSSIS] DEATHS      2 BOSTON    MA
4  TYPHOID FEVER [ENTERIC FEVER] DEATHS      2 BOSTON    MA
5                     DIPHTHERIA DEATHS      5 BOSTON    MA
6     WHOOPING COUGH [PERTUSSIS] DEATHS      2 BOSTON    MA
7  TYPHOID FEVER [ENTERIC FEVER] DEATHS      5 BOSTON    MA
8                     DIPHTHERIA DEATHS      8 BOSTON    MA
9  TYPHOID FEVER [ENTERIC FEVER] DEATHS      2 BOSTON    MA
10                 SCARLET FEVER DEATHS      1 BOSTON    MA
11                    DIPHTHERIA DEATHS      7 BOSTON    MA
12    WHOOPING COUGH [PERTUSSIS] DEATHS      3 BOSTON    MA
13 TYPHOID FEVER [ENTERIC FEVER] DEATHS      5 BOSTON    MA
14                    DIPHTHERIA DEATHS      7 BOSTON    MA
15    WHOOPING COUGH [PERTUSSIS] DEATHS      1 BOSTON    MA
16 TYPHOID FEVER [ENTERIC FEVER] DEATHS      7 BOSTON    MA
17                    DIPHTHERIA DEATHS      4 BOSTON    MA
18 TYPHOID FEVER [ENTERIC FEVER] DEATHS      2 BOSTON    MA
19                 SCARLET FEVER DEATHS      1 BOSTON    MA
20                    DIPHTHERIA DEATHS      5 BOSTON    MA
21 TYPHOID FEVER [ENTERIC FEVER] DEATHS      3 BOSTON    MA
22                    DIPHTHERIA DEATHS      5 BOSTON    MA
23    WHOOPING COUGH [PERTUSSIS] DEATHS      4 BOSTON    MA
24                  TYPHUS FEVER DEATHS      1 BOSTON    MA
25 TYPHOID FEVER [ENTERIC FEVER] DEATHS      6 BOSTON    MA
26                    DIPHTHERIA DEATHS      7 BOSTON    MA
27 TYPHOID FEVER [ENTERIC FEVER] DEATHS      6 BOSTON    MA
28                    DIPHTHERIA DEATHS     11 BOSTON    MA
29 TYPHOID FEVER [ENTERIC FEVER] DEATHS     10 BOSTON    MA
30                    DIPHTHERIA DEATHS      9 BOSTON    MA
31                       MEASLES DEATHS      1 BOSTON    MA
32 TYPHOID FEVER [ENTERIC FEVER] DEATHS     11 BOSTON    MA
33                    DIPHTHERIA DEATHS     15 BOSTON    MA
34    WHOOPING COUGH [PERTUSSIS] DEATHS      4 BOSTON    MA
35 TYPHOID FEVER [ENTERIC FEVER] DEATHS     11 BOSTON    MA
36                 SCARLET FEVER DEATHS      1 BOSTON    MA
37                    DIPHTHERIA DEATHS     11 BOSTON    MA
38    WHOOPING COUGH [PERTUSSIS] DEATHS      1 BOSTON    MA
39 TYPHOID FEVER [ENTERIC FEVER] DEATHS      8 BOSTON    MA
40                 SCARLET FEVER DEATHS      1 BOSTON    MA
41                    DIPHTHERIA DEATHS     13 BOSTON    MA
42                       MEASLES DEATHS      1 BOSTON    MA
43    WHOOPING COUGH [PERTUSSIS] DEATHS      1 BOSTON    MA
44 TYPHOID FEVER [ENTERIC FEVER] DEATHS      8 BOSTON    MA
45                    DIPHTHERIA DEATHS      7 BOSTON    MA
   Year Month Day
1  1888     7  22
2  1888     7  22
3  1888     7  22
4  1888     7  29
5  1888     7  29
6  1888     7  29
7  1888     8   5
8  1888     8   5
9  1888     8  12
10 1888     8  12
11 1888     8  12
12 1888     8  12
13 1888     8  19
14 1888     8  19
15 1888     8  19
16 1888     8  26
17 1888     8  26
18 1888     9   2
19 1888     9   2
20 1888     9   2
21 1888     9   9
22 1888     9   9
23 1888     9   9
24 1888     9  16
25 1888     9  16
26 1888     9  16
27 1888     9  23
28 1888     9  23
29 1888     9  30
30 1888     9  30
31 1888     9  30
32 1888    10   7
33 1888    10   7
34 1888    10   7
35 1888    10  14
36 1888    10  14
37 1888    10  14
38 1888    10  14
39 1888    10  21
40 1888    10  21
41 1888    10  21
42 1888    10  21
43 1888    10  21
44 1888    10  28
45 1888    10  28

Part 1 Exercises

Re-familarize yourselves with R by doing the following

Download and read in epi information from WorEpi.tab.txt
Use the View(), str(), and summary() functions to examine the epi data set from Worcester
Use the min(), max(), and table() functions to learn about the data in each column
Use some logic statements to look for data for only specific years or only for specific diseases

Part 2: More Logic and Condensing Data

More logic operator to combine several logic tests at once

operator	meaning
&	and
\|	or

You can use these two operators to do several logic tests at once. For example to look at data from only 1890’s we need to access the rows that are greater than or equal to 1890 and that are less than 1900

head(bosEpi$Year >= 1890 & bosEpi$Year < 1900, n = 100)

  [1] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
 [10] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
 [19] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
 [28] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
 [37] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
 [46] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
 [55] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
 [64] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE
 [73] FALSE FALSE FALSE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [82]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
 [91]  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
[100]  TRUE

Or looking only at deaths from diphtheria

head(bosEpi$disease == "DIPHTHERIA" & bosEpi$event == "DEATHS" , n = 100)

  [1] FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE FALSE
 [10] FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE FALSE
 [19] FALSE  TRUE FALSE  TRUE FALSE FALSE FALSE  TRUE FALSE
 [28]  TRUE FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE FALSE
 [37]  TRUE FALSE FALSE FALSE  TRUE FALSE FALSE FALSE  TRUE
 [46] FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE FALSE  TRUE
 [55] FALSE  TRUE FALSE FALSE  TRUE FALSE  TRUE FALSE FALSE
 [64]  TRUE FALSE  TRUE FALSE  TRUE FALSE FALSE  TRUE FALSE
 [73] FALSE FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE
 [82] FALSE  TRUE FALSE  TRUE FALSE FALSE FALSE  TRUE FALSE
 [91] FALSE  TRUE FALSE  TRUE FALSE  TRUE FALSE FALSE  TRUE
[100] FALSE

Looking only at rows for diphtheria or scarlet fever

head(bosEpi$disease == "DIPHTHERIA" | bosEpi$disease == "SCARLET FEVER" , n = 100)

  [1] FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE FALSE
 [10]  TRUE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE FALSE
 [19]  TRUE  TRUE FALSE  TRUE FALSE FALSE FALSE  TRUE FALSE
 [28]  TRUE FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE
 [37]  TRUE FALSE FALSE  TRUE  TRUE FALSE FALSE FALSE  TRUE
 [46] FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE  TRUE
 [55] FALSE  TRUE FALSE FALSE  TRUE FALSE  TRUE FALSE FALSE
 [64]  TRUE  TRUE  TRUE FALSE  TRUE FALSE FALSE  TRUE FALSE
 [73] FALSE  TRUE  TRUE FALSE  TRUE  TRUE FALSE  TRUE  TRUE
 [82] FALSE  TRUE  TRUE  TRUE FALSE FALSE  TRUE  TRUE FALSE
 [91] FALSE  TRUE FALSE  TRUE  TRUE  TRUE FALSE FALSE  TRUE
[100] FALSE

You can combine even further with () to group logic statements

head((bosEpi$disease == "DIPHTHERIA" | bosEpi$disease == "SCARLET FEVER")
     & bosEpi$event == "DEATHS", n = 100)

  [1] FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE FALSE
 [10]  TRUE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE FALSE
 [19]  TRUE  TRUE FALSE  TRUE FALSE FALSE FALSE  TRUE FALSE
 [28]  TRUE FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE
 [37]  TRUE FALSE FALSE  TRUE  TRUE FALSE FALSE FALSE  TRUE
 [46] FALSE  TRUE FALSE FALSE  TRUE FALSE FALSE  TRUE  TRUE
 [55] FALSE  TRUE FALSE FALSE  TRUE FALSE  TRUE FALSE FALSE
 [64]  TRUE  TRUE  TRUE FALSE  TRUE FALSE FALSE  TRUE FALSE
 [73] FALSE  TRUE  TRUE FALSE  TRUE  TRUE FALSE  TRUE  TRUE
 [82] FALSE  TRUE  TRUE  TRUE FALSE FALSE  TRUE  TRUE FALSE
 [91] FALSE  TRUE FALSE  TRUE  TRUE  TRUE FALSE FALSE  TRUE
[100] FALSE

As before you can use these logic statements to select only rows with specific information For example you can take the rows for DIPHTHERIA or SCARLET FEVER and resulted in DEATHS

dipScarDeaths = bosEpi[(bosEpi$disease == "DIPHTHERIA" | 
                         bosEpi$disease == "SCARLET FEVER") &
                         bosEpi$event == "DEATHS",]

head(dipScarDeaths)

         disease  event number    loc state Year Month Day
2     DIPHTHERIA DEATHS      7 BOSTON    MA 1888     7  22
5     DIPHTHERIA DEATHS      5 BOSTON    MA 1888     7  29
8     DIPHTHERIA DEATHS      8 BOSTON    MA 1888     8   5
10 SCARLET FEVER DEATHS      1 BOSTON    MA 1888     8  12
11    DIPHTHERIA DEATHS      7 BOSTON    MA 1888     8  12
14    DIPHTHERIA DEATHS      7 BOSTON    MA 1888     8  19

Now it might make more sense to look at the data summed over the years rather than per day per month. One way to do this would be with logic and using the sum() function.
For example to get the total number of deaths from diphtheria in the year 1890

sum(dipScarDeaths[dipScarDeaths$disease == "DIPHTHERIA" &
                    dipScarDeaths$Year == 1890 , ]$number)

[1] 401

The total number of deaths from diphtheria in the year 1891

sum(dipScarDeaths[dipScarDeaths$disease == "DIPHTHERIA" &  dipScarDeaths$Year == 1891, ]$number)

[1] 233

Now you can see how this can get very time consuming and tedious and luckily R has a function called aggregate() to do all these sums calculations at once and gives output in a data.frame for us. This function takes three important pieces of information.

The first is relationship that you want to examine that you let R know by using column names and the ~ symbol (which means depends on) and the * which in this case means by, in this case we want to see how the number column depends on the disease and Year columns.
The second is the data.frame that contains your data
The third is the calculation you want to do, in this case sum

dipScarDeathsSum = aggregate(number~disease*Year, dipScarDeaths, sum)

head(dipScarDeathsSum)

        disease Year number
1    DIPHTHERIA 1888    121
2 SCARLET FEVER 1888      4
3    DIPHTHERIA 1889    128
4 SCARLET FEVER 1889      3
5    DIPHTHERIA 1890    401
6 SCARLET FEVER 1890     42

The most confusing part of this call is probably the first one argument number~disease*Year since it looks a little funky but how it should be interpreted is examine the number column depending on the disease and Year column. So if we wanted to examine the data even more specifically we could add on the Month column like so

dipScarDeathsSumByMonth = aggregate(number~disease*Month*Year, dipScarDeaths, sum)

head(dipScarDeathsSumByMonth)

        disease Month Year number
1    DIPHTHERIA     7 1888     12
2    DIPHTHERIA     8 1888     26
3 SCARLET FEVER     8 1888      1
4    DIPHTHERIA     9 1888     37
5 SCARLET FEVER     9 1888      1
6    DIPHTHERIA    10 1888     46

Also aggregate() can take many different functions to get information about data, for example we could look at the mean deaths per month

dipScarDeathsMeanPerMonth = aggregate(number~disease*Month, dipScarDeaths, mean)

dipScarDeathsMeanPerMonth

         disease Month   number
1     DIPHTHERIA     1 6.547945
2  SCARLET FEVER     1 2.899160
3     DIPHTHERIA     2 6.106870
4  SCARLET FEVER     2 2.685714
5     DIPHTHERIA     3 5.379310
6  SCARLET FEVER     3 2.852174
7     DIPHTHERIA     4 5.369863
8  SCARLET FEVER     4 3.099099
9     DIPHTHERIA     5 5.097902
10 SCARLET FEVER     5 2.925620
11    DIPHTHERIA     6 4.468085
12 SCARLET FEVER     6 2.459184
13    DIPHTHERIA     7 4.043165
14 SCARLET FEVER     7 2.047059
15    DIPHTHERIA     8 4.015038
16 SCARLET FEVER     8 1.968254
17    DIPHTHERIA     9 4.563910
18 SCARLET FEVER     9 1.616667
19    DIPHTHERIA    10 6.147887
20 SCARLET FEVER    10 1.926471
21    DIPHTHERIA    11 6.914894
22 SCARLET FEVER    11 2.298701
23    DIPHTHERIA    12 7.171053
24 SCARLET FEVER    12 2.638889

To see more details about aggregate use the help function

help(aggregate)

Part 2 Exercises

Now examine the Worcester epi data set by using | and & and the aggregate function

Try summing up the numbers of deaths by looking at each disease separately or by comparing diseases side by side.
Try seeing the average number of deaths per month for each disease
Try looking at specific years to see how the number of deaths differ per month by comparing all disease or try selecting only specific diseases to look at
Look at the 1920’s to see where the most deaths were coming from

Part 3: Customizing Plots

Now we will now plot the data from the previous parts and learn how to customize the plot by changing

x limits using xlim=c()
y limits using ylim=c()
x label using xlab=
y label using ylab=
title using main=
color using col=
line thickness using lwd=
plot type using ty=
type of frame using bty=
Adding a legend using legend()

For example lots plot some data about the number of deaths caused by diphtheria and scarlet fever in Boston

First lets extract the data and sum it again like before

dipScarDeaths = bosEpi[(bosEpi$disease == "DIPHTHERIA" | bosEpi$disease == "SCARLET FEVER") & bosEpi$event == "DEATHS",]

dipScarDeathsSum = aggregate(number~disease*Year, dipScarDeaths, sum)
head(dipScarDeathsSum)

        disease Year number
1    DIPHTHERIA 1888    121
2 SCARLET FEVER 1888      4
3    DIPHTHERIA 1889    128
4 SCARLET FEVER 1889      3
5    DIPHTHERIA 1890    401
6 SCARLET FEVER 1890     42

Now lets plot the number of deaths for diphtheria by calling the plot() function and telling it to plot years on the x axis and number on the y axis

plot(dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$Year,
     dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$number )

The default plot by calling plot() is a scatter plot which might not be the best way to represent this data so we can use type= to change the type of plot, some options are “l” for line plot, “b” for both a line and scatter plot, and “h” for bars similar to a histogram, lets go with the last one

plot(dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$Year,
     dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$number,
     type = "h")

Now the default labels look all messed up so lets change them using xlab= and ylab= and we can add a title as well by using main= (which stands for main title)

plot(dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$Year,
     dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$number,
     type = "h", xlab = "Year", ylab = "Number of Deaths",
     main = "Deaths by SCARLET FEVER")

Now lets takes away the frame because it doesn’t add much using bty= (which stands for box type) and make the lines thicker as well by using lwd= (which stands for line width)

plot(dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$Year,
     dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$number,
     type = "h", xlab = "Year", ylab = "Number of Deaths", 
     main = "Deaths by SCARLET FEVER", bty = "none", lwd = 2)

Lets add on the data points from the deaths of scarlet fever by using the points() function, which is just like the plot() function but puts it’s points over the current plot rather than creating a new plot and since we are just adding the points we just need to tell it the type and the line width. Also since we want to see the two data points next to each other, let’s offset the year by 0.5 so the points end up next to each other. Also let’s add color by using the col= function

plot(dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$Year,
     dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$number,
     type = "h", xlab = "Year", ylab = "Number of Deaths",
     main = "Deaths by Diphtheria vs Deaths by Scarlet Fever",
     bty = "none", lwd = 2, col = "red")  

points(dipScarDeathsSum[dipScarDeathsSum$disease == "DIPHTHERIA",]$Year + 0.5,
       dipScarDeathsSum[dipScarDeathsSum$disease == "DIPHTHERIA",]$number,
       type = "h", lwd = 2, col = "darkblue")

It looks like the data points for deaths by Diphtheria are being cut off because the default limites were set by the max number of Scarlet Fever deaths so lets change the xlimits and ylimits by xlim=c() and ylim=c() to tell the plot to span certain limits. We can also take advantage of the R function range() For example if we call range() on the numbers column we can get the max number and min numbers

range(dipScarDeathsSum$number)

[1]   3 837

plot(dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$Year,
     dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$number,
     type = "h", xlab = "Year", ylab = "Number of Deaths",
     main = "Deaths by Diphtheria vs Deaths by Scarlet Fever",
     bty = "none", lwd = 2, xlim = range(dipScarDeathsSum$Year),
     ylim = range(dipScarDeathsSum$number), col= "red")  

points(dipScarDeathsSum[dipScarDeathsSum$disease == "DIPHTHERIA",]$Year + 0.5,
       dipScarDeathsSum[dipScarDeathsSum$disease == "DIPHTHERIA",]$number, 
       type = "h", lwd = 2, col = "darkblue")

Since there isn’t much data after the 1920 it might be more interesting to look at only the data through that time point so we can see the data in more detail, lets change the xlim=c() to the change the year range

plot(dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$Year,
     dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$number
     , type = "h", xlab = "Year", ylab = "Number of Deaths", 
     main = "Deaths by Diphtheria vs Deaths by Scarlet Fever", 
     bty = "none", lwd = 2, xlim =c(1880, 1930), 
     ylim = range(dipScarDeathsSum$number), col= "red") 

points(dipScarDeathsSum[dipScarDeathsSum$disease == "DIPHTHERIA",]$Year + 0.5, 
       dipScarDeathsSum[dipScarDeathsSum$disease == "DIPHTHERIA",]$number,
       type = "h", lwd = 2, col = "darkblue")

We can also add a legend

plot(dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$Year,
     dipScarDeathsSum[dipScarDeathsSum$disease == "SCARLET FEVER",]$number,
     type = "h", xlab = "Year", ylab = "Number of Deaths", 
     main = "Deaths by Diphtheria vs Deaths by Scarlet Fever",
     bty = "none", lwd = 2, xlim =c(1880, 1930), 
     ylim = range(dipScarDeathsSum$number), col= "red")  

points(dipScarDeathsSum[dipScarDeathsSum$disease == "DIPHTHERIA",]$Year + 0.5,
       dipScarDeathsSum[dipScarDeathsSum$disease == "DIPHTHERIA",]$number,
       type = "h", lwd = 2, col = "darkblue") 

legend("topright", legend = c("SCARLET FEVER", "DIPHTHERIA"),
       lwd = 2, col = c("red", "darkblue"))

Part 3 Exercises

Try customizing some plots for the diseases found in the Worcester Epi Dataset
For example trying to look how deaths have decreased over time for some diseases like measles (though maybe on the raise again), smallpox, etc
Try creating a plot comparing deaths between SCARLET FEVER and TYPHOID FEVER [ENTERIC FEVER]
Try ploting a couple of the big diseases in the 1920’s

Part 4 Installing and loading R packages

Now we can use the colors provided by R but we can also use an R package called RColorBrewer or colorspace to provide us with even more colors but first we have to learn how to download the packages which can be done using install.packages() and library() or we can use RStudio to download them

Using install.packages() and library()

#install package
install.packages("RColorBrewer")
#Once package is installed to actually use it you need to load it
library("RColorBrewer")

Or we can use RStudio to manually install the packages
In the bottom right corner of RStudio click the Packages tab and then the install icon

Type in RColorBrewer

To Load click the check box next to RColorBrewer

Data Visualization in R: Session 3

October 1, 2014