If you can see this message: congratulations! It means that you have
sucessfully installed R and RStudio on your
computer and that you are ready to dive into building those basic skills
that will lead you to being able to manipulate data frames, tables of
data where each row represents an observation (e.g., a person) and each
column represents a measurement (e.g., the height or weight of that
person).
Before we start, note that when you insert code into each “code chunk” below (the sections delimited with ```{r} and ```), you can test it in a number of ways:
Each column of a data frame is itself a vector. So let’s start with
some basic vector manipulation. Use the c() function to
define a vector x with four values: 1, 2, 1, and 8.
(You should replace the “# FILL ME IN” statement below with your
answer.) Note that vectors are homogeneous (all of the same type),
with the most common types being double (or
numeric), character, and logical.
The vector you define has type double. Check this by typing
typeof(x) and noting the output. Then type
x[1]. What do you see? (Indicate the answer by, on a new
line in the code chunk, typing one number symbol # [which denotes a
comment] and then afterwards typing your answer in words [and/or
numbers].)
x <- c(1,2,1,8)
typeof(x)## [1] "double"x[1]## [1] 1# type of x gives the data type of collection x. And x[1] gives the output of the first element of collection/dataframe x.If the value(s) inside the square bracket is/are numeric, then
that/those elements of the vector are displayed. (Note: R
counts from 1, not 0.) If the value(s) are logical, then only those
elements with value TRUE are displayed. This will make more
sense below.
Now define a vector y with four values: 2, 2, 5, 8. Then
add x and y, and multiply x and
y. Note that the following operators are using to carry out
basic math in R:
| Operation | Description |
|---|---|
| + | addition |
| - | subtraction |
| * | multiplication |
| / | division |
| ^ | exponentiation |
| %% | modulus (i.e., remainder) |
| %/% | division with (floored) integer round-off |
y <- c(2,2,5,8)
x + y## [1] 3 4 6 16x * y## [1] 2 4 5 64# FILL ME INWhat you should observe are vectors with four numbers each. Note that
R did not require you to loop over the vector indices,
i.e., R did not make you add x[1] and
y[1] first, then x[2] and y[2],
etc. R made things easy, by utilizing
vectorization: it takes care of entire vectors at once, without
explicit loops needing to be defined by you.
Now redefine the vector x to be of length 500, with all
the elements being 0. You don’t want to do this using the
c() function! (Look back at section 2.1 of
cmu-intro-r.github.io for alternative ways to define
vectors.)
# FILL ME INx <- rep(0 , 500)
Let’s define a random vector of integers with the
sample() function. The information that you pass into a
function is called an argument, and R functions sometimes
can have many arguments. Let’s look at the help page for
sample(), which you can bring up by typing
?sample in the console, or going to the Help
pane and typing sample in the search bar.
Usage
sample(x, size, replace = FALSE, prob = NULL)(What I just typed is an example of a verbatim block: it doesn’t
execute as R code.) What we see is that
sample() has four arguments. Two of them,
replace and prob, have default
values…so if you are happy with the defaults, you need not specify
values for these arguments at all. So you just need to specify, at a
minimum, two arguments: x, which is either a number or a
vector from which to sample data, and size, which is the
number of data to sample. If you do this
x <- sample(10,5)you are telling R to sample five numbers between 1 and
10 (inclusive), with all the numbers being different (because
replace=FALSE), and to save the numbers as the vector
x. If you do this
x <- sample(40:50,5)you are telling R to sample five different numbers
between 40 and 50 (inclusive). And if you do this
x <- sample(3,10,replace=TRUE)
print(x)you are telling R to sample ten numbers between 1 and 3
(inclusive), and repetition is allowed. (We call this “sampling with
replacement.”) Etc. Now, sample 100 numbers between 1 and 100
(inclusive) with replacement, and save the output as the vector
x. How many unique integers are there in x?
Use handy vector functions (see section 2.3) to get a concise answer: do
not print out x and count by eye! (If you need help, call a
TA or me over, or come to office hours.)
x <- sample(3,10,replace=TRUE)
print(x)## [1] 2 3 3 2 3 3 1 1 1 3x <- sample(100,100,replace = TRUE)
length(unique(x))## [1] 66Relational operators are binary operators of the form “variable
operator value,” e.g., x < 0. The six basic relational
operators are ==, !=, <,
>, <=, and >= (for
“equals,” “not equals,” “less than,” “greater than,” “less than or
equals,” and “greater than or equals.”) Relational operators return a
vector of logicals, meaning a vector of TRUE and
FALSE values. Below, redefine x to be the
vector with elements 1, 2, 1, and 8, and then display the output for
x == 1 and x > 3.
# FILL ME IN
x <- c(1,2,1,8)
print(x == 1)## [1] TRUE FALSE TRUE FALSEprint(x > 3)## [1] FALSE FALSE FALSE TRUEApply the sum() function with input x == 1.
Does the output make sense to you?
# FILL ME INsum(x == 1) ## Question 7
Relational operators may be combined with & (logical
AND) or | (logical OR). Below, display the output for
x < 2 | x > 5.
# FILL ME IN
print(x<2 | x>5)## [1] TRUE FALSE TRUE TRUEA reason to learn relational operators is that they underpin the
manipulation of vectors (and thus underpin the manipulation of, e.g.,
rows or columns of data frames). To display a subset of values of the
vector x, you can for instance type x[...],
where you would replace ... with a relational operator.
What happens when you type x[x==1]?
# FILL ME IN
x[x==1]## [1] 1 1Some last things to do for now: apply the length()
function to x, apply the sort() function to
x, apply the sort() function to x
with the additional argument decreasing=TRUE, apply the
unique() function to x, and apply the
table() function to x. (You may have done some
similar things above when we told you to solve certain problems with
handy vector functions.) Build intuition about what each does. (Note
that table() is a handy function for doing exploratory data
analysis of categorical variables.)
# FILL ME IN
length(x)## [1] 4sort(x, decreasing = TRUE)## [1] 8 2 1 1unique(x)## [1] 1 2 8table(x)## x
## 1 2 8
## 2 1 1(Looking ahead.) A list in R is a collection of
vectors. Define a list below using list(), with the first
argument being a defined vector with name x and values 1
and 2, and the second argument being a defined vector with name
y and values “a”, “b”, and “c”. (Note: your arguments won’t
look like z <- c(TRUE,FALSE) but more like
"z"=c(TRUE,FALSE)). Display the list.
# FILL ME IN
z <- list(x=1:2,y=c("a","b","c"))
print(z)## $x
## [1] 1 2
##
## $y
## [1] "a" "b" "c"The individual entries of a list are vectors, which are homogeneous, but the entries may each be of different type. A list whose entries are all of the same length is a data frame (i.e., a structured data table).
Download simple.txt from the Week 01 module page. Apply
an appropriate function to read the file’s contents into R.
(Are the data separated by commas or spaces? Note that when you click on
the file link, the file’s contexts should appear in Canvas,
along with a download link.) Show the names of the columns. Make sure
the names are correct, and that there are eight columns. (Use
ncol() to determine the number of columns.) Note: you
should consider including the argument
stringsAsFactors=FALSE.
# FILL ME IN
df = read.csv("simple.txt",header = TRUE, sep = " " , stringsAsFactors = FALSE)
ncol(df)## [1] 8colnames(df)## [1] "name" "u" "g" "r" "i" "z" "y"
## [8] "redshift"Read in the data file from Question 11 but skip the header. Display
the names that R gives to the columns. Note that the
skip argument can be very useful for skipping over any
metadata that may be present in your file.
# FILL ME IN
df = read.csv("simple.txt",header = FALSE, sep = " " , skip = 1, stringsAsFactors = FALSE)
print(df)## V1 V2 V3 V4 V5 V6 V7 V8
## 1 galaxy.A 17.8313 16.9077 16.4431 16.2099 16.0613 15.8732 0.038356
## 2 galaxy.B 19.0731 17.7448 16.9789 16.5288 16.2551 15.9531 0.058309
## 3 galaxy.C 21.6380 21.0106 20.8286 20.6283 20.6552 20.5280 0.063701
## 4 galaxy.D 20.5474 19.5542 19.2387 19.0568 19.0887 18.9865 0.059006
## 5 galaxy.E 21.2378 20.6876 20.5661 20.4371 20.4799 20.4503 0.063202
## 6 galaxy.F 22.4627 21.4597 21.0484 20.8274 20.7639 20.6385 0.057773
## 7 galaxy.G 23.8221 22.8950 22.5779 22.3543 22.3225 22.2038 0.061548
## 8 galaxy.H 23.0491 22.1536 21.8791 21.6889 21.7044 21.6381 0.063769
## 9 galaxy.I 23.6742 23.0346 22.7857 22.6116 22.5813 22.5462 0.061427Read in the data file from Question 11 but only read in the first
four lines, while retaining the header. Print the data frame. (Note that
in general, you can use skip and nrows to zero
in on portions of a text file where the data actually reside.)
# FILL ME IN
df = read.csv("simple.txt",header = TRUE, sep = " " , nrows = 4, stringsAsFactors = FALSE)
print(df)## name u g r i z y redshift
## 1 galaxy.A 17.8313 16.9077 16.4431 16.2099 16.0613 15.8732 0.038356
## 2 galaxy.B 19.0731 17.7448 16.9789 16.5288 16.2551 15.9531 0.058309
## 3 galaxy.C 21.6380 21.0106 20.8286 20.6283 20.6552 20.5280 0.063701
## 4 galaxy.D 20.5474 19.5542 19.2387 19.0568 19.0887 18.9865 0.059006Download planets_small.csv from the Week 01 module page.
Apply an appropriate function to read the file’s contents into
R. Note that here, you have one column that should be
rendered as character strings (pl_hostname, the first
column), while the rest should be rendered as factor variables. Thus you
do not want to use the stringsAsFactors argument here, as
it is too coarse. You need to explicitly specify the types of each
column, using the functional argument colClasses. After
reading the data in, pass your data frame into the function
summary() and see what happens.
# FILL ME IN
df = read.csv("planets_small.csv",header = TRUE, sep = "," , colClasses=c("character","factor","character","integer"))
print(df)## pl_hostname pl_letter pl_discmethod pl_pnum
## 1 11 Com b Radial Velocity 1
## 2 11 UMi b Radial Velocity 1
## 3 14 And b Radial Velocity 1
## 4 14 Her b Radial Velocity 1
## 5 16 Cyg B b Radial Velocity 1
## 6 18 Del b Radial Velocity 1
## 7 1RXS J160929.1-210524 b Imaging 1
## 8 24 Sex b Radial Velocity 2
## 9 24 Sex c Radial Velocity 2
## 10 2MASS J01225093-2439505 b Imaging 1summary(df)## pl_hostname pl_letter pl_discmethod pl_pnum
## Length:10 b:9 Length:10 Min. :1.0
## Class :character c:1 Class :character 1st Qu.:1.0
## Mode :character Mode :character Median :1.0
## Mean :1.2
## 3rd Qu.:1.0
## Max. :2.0Download emline.csv from the Week 01 module page. Apply
an appropriate function to read the file’s contents into R.
When you are done, show the mean and median values of the
sfr column. Hint: if they are wildly different, you may
need to adjust how you read in the data. Hint: look for numbers that
represent missing data, and use an appropriate argument to tell
R that those numbers should be converted to
NA. Note that if are converted to NA, you will
have to pass an additional argument into the mean() and
median() functions: na.rm=TRUE, which tells
R to ignore the NAs in computations.
# FILL ME IN
df = read.csv("emline.csv",header = TRUE, sep = ",")
print(df)## O_II_3729 H_gamma H_beta O_III_4959 O_III_5007 N_II_6548 H_ALPHA N_II_6584
## 1 1.5364 1.7004 0.3709 -0.0549 0.6216 0.4846 1.2458 0.5687
## 2 9.4411 2.7609 4.0603 0.5418 0.8290 0.7808 11.2330 4.2059
## 3 17.3146 3.3528 6.3293 0.7573 4.7597 1.0888 21.6571 3.5426
## 4 8.8070 -0.2765 1.4274 1.0690 1.7463 0.5824 7.5148 2.0032
## 5 17.3470 3.8775 2.6622 0.5714 3.2273 1.1112 13.9770 4.1352
## 6 37.8111 4.4961 10.0149 3.2715 11.7268 2.2371 39.9036 8.3394
## 7 55.3728 7.7921 13.7503 9.4787 28.0838 1.3237 38.2856 3.9767
## 8 24.0244 4.6931 8.4271 1.9320 6.4095 2.6659 35.1531 10.7334
## 9 31.8309 3.2743 7.0494 3.7526 12.1558 2.3935 31.8069 5.3516
## S_II_6717 S_II_6731 mass sfr
## 1 -0.4526 1.0754 9.9484 -9999.0000
## 2 1.8614 2.6442 9.9976 -9999.0000
## 3 2.5171 3.1083 9.8058 -0.7249
## 4 2.7998 0.7398 9.8352 -9999.0000
## 5 4.6136 2.3890 9.7140 -0.6436
## 6 9.9219 6.2621 9.6657 -0.4929
## 7 3.2716 4.6510 8.9972 -9999.0000
## 8 9.7281 5.5440 9.8578 -0.5099
## 9 8.5123 10.3885 9.8120 -1.1993mean(df$sfr, trim = 0 , na.rm = FALSE)## [1] -4444.397median(df$sfr, trim = 0 , na.rm = FALSE)## [1] -1.1993