R Basics / Course Business l Well be using a sample dataset in class - - PowerPoint PPT Presentation
R Basics / Course Business l Well be using a sample dataset in class today: l CourseWeb: Course Documents Sample Data Week 2 l Can download to your computer before class l Thanks for answering CourseWeb background survey! l If sitting in
l We’ll be using a sample dataset in class today:
l CourseWeb: Course Documents à Sample Data
à Week 2
l Can download to your computer before class
l Thanks for answering CourseWeb background
survey!
l If sitting in on the course, e-mail me so I can
add you to CourseWeb
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
l Simplest way to interact with R is by
typing in commands at the > prompt:
l Typing in a simple calculation shows us
the result:
l 608 + 28
l What’s 11527 minus 283? l Some more examples:
l 400 / 65
(division)
l 2 * 4 (multiplication) l 5 ^ 2 (exponentiation)
l More complex calculations can be done
with functions:
l sqrt(64) l Can often read these
left to right (“square root of 64”)
l What do you think
this means?
l abs(-7)
What the function is (square root) In parenthesis: What we want to perform the function on
l Some functions have settings
(“arguments”) that we can adjust:
l round(3.14)
decimal places)
l round(3.14, digits=1)
l We can use multiple functions in a row,
l Don't get scared when you see multiple
parentheses!
the middle
l When we want to use multiple numbers,
we concatenate them
l c(2,6,16)
l Sometimes a computation requires
multiple numbers
l Also a quick way to do the same thing to
multiple different numbers:
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
Course Documents: Sample Data: Week 2
l Reading plausible versus implausible
sentences
l “Scott chopped the carrots with a knife.”
“Scott chopped the carrots with a spoon.”
Measure reading time
word
Note: Simulated data; not a real experiment.
Course Documents: Sample Data: Week 2
l Reading plausible versus implausible
sentences
l Reading time on critical word l 36 subjects l Each subject sees 30 items (sentences):
half plausible, half implausible
l Interested in changes over time, so we’ll
track number of trials remaining (29 vs 28 vs 27 vs 26…)
l Make sure you have the dataset at this
point if you want to follow along: Course Documents à Sample Data à Week 2
l Navigate to the
folder in lower-right
l More ->
Set as Working Directory
l Open a “comma-separated value” file:
Name of the “dataframe” we’re creating (whatever we want to call this dataset) read.csv is the function name File name
l Read in a “comma-separated value” file:
('/Users/scottfraundorf/Desktop/week2.csv') Name of the “dataframe” we’re creating (whatever we want to call this dataset) read.csv is the function name Folder & file name
full folder & filename
l A “big picture” of the dataset:
l summary(experiment)
l summary() is a very important function!
l Basic info & descriptive statistics l Check to make sure the data are correct
l A “big picture” of the dataset:
l summary(experiment)
l We can use $ to refer to a specific
column/variable in our dataset:
l summary(experiment$ItemName)
l Let’s look at the data!
l experiment
l Ack! That’s too much! How about just a
few rows?
l head(experiment) l head(experiment, n=10)
l Excel:
read.xls('/Users/scottfraundorf/De sktop/week2.xls')
l SPSS:
read.spss('/Users/scottfraundorf/D esktop/week2.spss', to.data.frame=TRUE)
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
l Save & reuse commands with a script
File -> New Document
l Run commands without typing them all
again
l R Studio:
l Code -> Run Region -> Run All: Run entire script l Code -> Run Line(s): Run just what you’ve
highlighted/selected
l R:
l Keyboard shortcut for this:
l Saves times when re-running analyses l Other advantages? l Some:
perform one analysis after another
l Add # before a line to make it a
comment
(or other readers)
line a comment
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
l Remember how we referred to a
particular variable in a dataframe?
l Combine that with functions:
l Or, for a categorical variable:
l We often want to look at a dependent variable
as a function of some independent variable(s)
experiment$Condition, mean)
l Try getting the mean RT for each item l How about the median RT for each subject? l To combine multiple results into one table,
“column bind” them with cbind():
l cbind(
tapply(experiment$RT, experiment$Condition, mean), tapply(experiment$RT, experiment$Condition, sd) )
l Can have 2-way tables...
list(experiment$Subject, experiment$Condition), mean)
l ...or more!
list(experiment$ItemName, experiment$Condition, experiment$TestingRoom), mean)
l Contingency tables for categorical
variables:
data=experiment)
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
l Often, we want to examine or use just
part of a dataframe
l Remember how we read our dataframe?
l Create a new dataframe that's just a
subset of experiment:
subset(experiment, RT < 2000)
Inclusion criterion: RT less than 2000 ms Original dataframe New dataframe name
l Try getting just the observations with
RTs 200 ms or more:
subset(experiment, RT >= 200)
l Why not just delete the bad RTs from the
spreadsheet?
l Easy to make a mistake / miss some of them l Faster to have the computer do it l We’d lose the original data l No documentation of how we subsetted the data
l What if we wanted only RTs between 200
and 2000 ms?
subset(experiment, RT >= 200)
subset(experiment.Temp, RT <= 2000)
l One step with & for AND:
RT >= 200 & RT <= 2000)
l What if we wanted only RTs between 200
and 2000 ms?
l One step with & for AND:
RT >= 200 & RT <= 2000)
l | means OR:
subset(experiment, RT < 200 | RT > 2000)
l Get a match / equals:
subset(experiment, TrialsRemaining == 0)
l Words/categorical variables need quotes:
subset(experiment, Condition=='Implausible')
l != means “not equal to”:
subset(experiment, Subject != 'S23')
Note DOUBLE equals sign Drops subject “S23”
l Sometimes our inclusion criteria aren't so
mathematical
l Suppose I just want the “Ducks” and
“Panther” items
l We can check against any arbitrary list:
subset(experiment, ItemName %in% c('Ducks', 'Panther'))
l Or, keep just things that aren't in a list:
<- subset(experiment, Subject %in% c('S10', 'S23') == FALSE)
l Note that these subsets are just creating
new dataframes in R
l If you want to save to a folder on your
computer, use write.csv():
l write.csv(experiment.BadRTsRemoved,
file='experiment_badremoved.csv')
l Summary
Greater than
Greater than or equal to
Less than
Less than or equal to
AND
OR
Equal to
Not equal to
Is this included in a list?
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
l Try getting the mean RT for each number
28 trials remaining, etc.)
l Try getting a subset of just the people in
TestingRoom 3
l Try getting the mean RT for each number
28 trials remaining, etc.)
l tapply(experiment$RT,
experiment$TrialsRemaining, mean)
l Try getting a subset of just the people in
TestingRoom 3
l experiment.Room3 <- subset(experiment,
TestingRoom == 3)
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
l Remember the pointing arrow used to
create dataframes and subsets?
l This is the assignment operator. It
saves results or values in a variable
you the current value:
l Assigning a new value overwrites the old
l We can use this to create new columns
in our dataframe:
every trial
l Or, compute a value for each row:
experiment$RT / 1000
specific trial and saves that into RTinSeconds
l We can use this to create new columns
in our dataframe:
every trial
l Another example:
30 - experiment$TrialsRemaining
(trial #1, trial #2, etc.) and saves the result into
SerialPosition
l ifelse(): Use a test to decide which of two
values to assign:
l experiment$Half <- ifelse(
experiment$SerialPosition <= 15, 1, 2)
l Possible to nest ifelse() if we need more
than 2 categories:
l experiment$Third <-
ifelse(experiment$SerialPosition <= 10, 1, ifelse(experiment$SerialPosition <= 20, 2, 3))
Function name If serial position IS <= 15… “Half” is 1 If it’s NOT, “Half” is 2
l Instead of specific numbers, can use
l experiment$RT.Fixed <- ifelse(
experiment$TestingRoom==2, experiment$RT + 100, experiment$RT)
l How can we check if this worked?
l head(experiment)
Fixed RTs are indeed 100 ms longer for TestingRoom 2
l Instead of specific numbers, can use
l experiment$RT.Fixed <- ifelse(
experiment$TestingRoom==2, experiment$RT + 100, experiment$RT)
l How can we check if this worked?
l tapply(experiment$RT,
experiment$TestingRoom, mean)
l tapply(experiment$RT.Fixed,
experiment$TestingRoom, mean)
l Only room 2 should be different
ifelse(experiment$SerialPosition <= 15, 1, 2)
l vs:
ifelse(experiment$SerialPosition <= (CriticalTrialsPerSubject / 2) , 1, 2)
to this script later
variable later in the script & this ensure it’s consistent
l It is also possible to delete a variable by
assigning it the value of NULL
don’t want to bother keeping TrialsRemaining any more
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
l So far, we’ve seen how to
existing dataframe
l What if we want to modify a dataframe by
adjusting some existing values?
l e.g., replace all RTs above 2000 ms with the
number 2000 (“fencing”)
l Creating a new subset won’t work because we
want to change the original dataframe
l Need a way to edit specific values
l Use square brackets [ ] to refer to specific
entries in a dataframe:
l Omit the row or column number to mean all
rows or all columns:
l Can also use column names:
l experiment[,'RT']
All rows in the RT column
l Remember c()? We can check multiple rows:
Row 3, all columns All rows in column 4
l We can look at rows or columns that meet
a specific criterion...
l Can use this as another way to subset:
experiment[experiment$RT > 200, ]
l But we can also set values this way
'RT'] <- 200
where RT < 200, and set the column RT to 200
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
l R treats continuous & categorical variables
differently:
l These are different data types:
categories (e.g., treatment vs. placebo)
l R's heuristic when reading in data:
factor
l Sometimes we need to correct this
l Create a new column that's the factor
(categorical) version of TestingRoom:
as.factor(experiment$TestingRoom)
l Or, just overwrite the old column:
as.factor(experiment$TestingRoom)
l When ifelse() results in words, R creates
a character variable rather than a factor
l Wrong:
ifelse(experiment$TestingRoom==3, 'My favorite room', 'Not favorite')
l Right:
as.factor(ifelse(experiment$TestingRo
favorite'))
l To change a factor to a number, need to
turn it into a character first:
as.numeric(as.character(experiment$ Age))
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
l We might have run into some problems
trying to change Age into a numerical variable...
l NA means “not available”...
l If we try to do computations on a set of
numbers where any of them is NA, we get
NA as a result...
l R wants you to think about how you want
to treat these missing values
l To ignore the NAs when doing a specific
computation, use na.rm=TRUE:
l To get a copy of the dataframe that excludes
all rows with an NA (in any column):
l Change NAs to something else with logical
indexing:
Age.Numeric)==TRUE, ]$Age.Numeric <- 23
l R commands & functions l Reading in data l Saving R scripts l Descriptive statistics l Subsetting data l Assigning new values l Referring to specific cells l Types & type conversion l NA values l Getting help
l Get help on a specific known function:
arguments
l Try to find a
function on a particular topic:
l Some built-in analyses:
l aov()
ANOVA
l lm()
Linear regression
l glm()
Generalized linear models (e.g., logistic)
l cor.test()
Pearson correlation
l t.test()
t-test
l Help function (?) will tell you about the
arguments to these particular functions
l Can use R for: