In the last post on R, we began to look at how to pull information from a data set. In this post, we will continue this by examining how to describe continuous variables. One of the first steps in any data analysis is to look at the descriptive statistics to determine if there are any problems, errors or issues with normality. Below is the code for the data frame we created.

> cars <- mtcars[c(1,2,9,10)]
> cars$am <- factor(cars$am, labels=c('auto', 'manual'))
> cars$gear <- ordered(cars$gear)

Finding the Mean, Median, Standard Deviation, Range, and Quartiles

The mean is useful to know as it gives you an idea of the centrality of the data. Finding the mean is simple and involves the use of the “mean” function. Keep in mind that there are four variables in our data frame which are mpg, cyl, am, and gear. Only ‘mpg’ has a mean because the other variables are all categorical (cyl, am, and gear). Below is the script for finding the mean of ‘mpg’ with the answer.

> mean(cars$mpg)
[1] 20.09062

The median is the number found exactly in the middle of a dataset. Below is the median of ‘mpg.’

> median(cars$mpg)
[1] 19.2

The answer above indicates that the median of ‘mpg’ is 19.2. This means half the values are above this number while half the values are below it.

Standard deviation is the average amount that a particular data point is different from the mean. For example, if the standard deviation is 3 and the mean is 10 this means that any given data point is either is between 7 and 13 or -3 to  +3 different from the mean. In other words, the standard deviation calculates the average amount of deviance from the mean. Below is a calculation of the standard deviation of the ‘mpg’ in the ‘cars’ data frame using ‘sd’ function.

> sd(cars$mpg)
[1] 6.026948

What this tells is that the average amount of deviance from the mean of 20.09 for ‘mpg’ in the ‘car’ data frame is 6.02. In simple terms, most of the data points are between 14.0 and 26.0 mpg.

Range is the highest and lowest number in a data set. It gives you an idea of the scope of a dataset. This can be calculated in R using the ‘range’ function. Below is the range for ‘mpg’ in the ‘cars’ data frame.

> range(cars$mpg)
[1] 10.4 33.9

These results mean that the lowest ‘mpg’ in the cars data frame is 10.4 while the highest is 33.9. There are no values lower or higher than these.

Lastly, quartiles tell breaks the data into several groups based on a percentage. For example, the 25th percentile gives you a number that tells that 75% of the numbers are higher than it and that 25% is lower. If there are 100 data points in a dataset and the number 25 is the 25th percentile, this means that 75% or 75 numbers are of greater value than 25 and that there are about 25 numbers lower than 25. Below is an example using the ‘mpg’ information from the ‘cars’ data frame using the ‘quantile’ function.

> quantile(cars$mpg)
    0%    25%    50%    75%   100% 
10.400 15.425 19.200 22.800 33.900

In this example above, 15.4 is the 25th percentile. This means that 75% of the values are above 15.4 while 25% are below it. In addition, you may have noticed that the 0% and the 100% are the same as the range. Furthermore, the 50% is the same as the median. In other words, calculating the quartiles gives you the range and median. Therefore, calculating the quartiles saves you time.

Conclusion

These are the basics of describing continuous variables. The next post on R will look at describing categorical variables.

In this post and future posts, we will work with actual data that is available in R to apply various skills. For now, we will work with the ‘mtcars’ data that is available in R. This dataset contains information about 32 different cars that were built in the 1970’s.

Often in research, the person who collects the data and the person who analyzes the data (the statistician) are different. This often leads to the statistician getting a lot of missing data that cannot be analyzed in its current state. This leads to the statistician having to clean the data in order to analyze it.

Initial Decisions

One of the first things a statistician does after he has received some data is to see how many different variables there are and perhaps decide if any can be converted to factors. In order to achieve these two goals we need to answer the following questions

1. How many variables are there? Many functions can answer this question
2. How many unique values does each variable have? This will tell us if the variable is a candidate for becoming a factor.

Below is the code for doing this with the ‘mtcar’ dataset.

> sapply(mtcars, function(x) length(unique(x)))
 mpg  cyl disp   hp drat   wt qsec   vs   am gear carb 
  25    3   27   22   22   29   30    2    2    3    6

Here is what we did

1. We used the ‘sapply’ function because we want to apply our function on the whole dataframe at once.
2. Inside the ‘sapply’ function we tell R that the dataset is ‘mtcars’
3. Next is the function we want to use. We tell R to use the function ‘x’ which is an anonymous function.
4. After indicating that the function is anonymous we tell R what is inside the anonymous function.
5. The anonymous function contains the length function with the unique function within it.
6. This means that we want the length (or number) of unique values in each variable.

We have the answers to our question

1. There are eleven variables in the ‘mtcars’ dataset as listed in the table above
2. The unique values are also listed in the table above.

A common rule of thumb for determining whether to convert a variable to a factor is that the variable has less then 10 unique values. Based on this rule, the cyl, vs, am, gear, and carb variables could be converted to factors. Converting to a factor makes a continuous variable a categorical one and opens up various analysis possibilities.

Preparing Data

Now that we have an idea of what are the characteristics of the dataset we have to decide the following…

1. What variables are we going to use
2. What type of analysis are we going to do with the variables we use.

For question 1 we are going to use the 1st (mpg), the 2nd (cyl), the 9th (am) and the 10th (gear) variables for our analysis. Then we are going to make the ‘am’ variable a factor. Lastly, we are going to put in order the values in the ‘gear’ variable. Below is the code

> cars <- mtcars[c(1,2,9,10)]
> cars$am <- factor(cars$am, labels=c('auto', 'manual'))
> cars$gear <- ordered(cars$gear)

Here is what we did

1. We created the variable ‘cars’ and assigned a subset of variables from the ‘mtcars’ dataset. In particular, we took the  1st (mpg), the 2nd (cyl), the 9th (am) and the 10th (gear) variables from ‘mtcars’ and saved them in ‘cars’
2. For the ‘am’ variable we converted it to a factor. Data points that were once 0 became ‘auto’ and data points that were once 1 became ‘manual’
3. We made the ‘gear’ variable an ordered factor this means that 5 is more than 4, 4 is more than 3, etc.

Our next post on R will focus on analyzing the ‘cars’ variable that we created.

The last R post focused on the use of the “for” loop. This option is powerful in repeating an action that cannot be calculated in a vector. However, there are some drawbacks to ‘for’loops that are highly technical and hard to explain to beginners. The problems have something to do with strange results in the workspace and environment.

To deal with the complex problems with ‘for’ loops have the alternative approach of using functions from the apply family. The functions in the apply family provide the same results as a ‘for’ loop without the occasional problem. There are three functions in the apply family and they are

• apply
• sapply
• lapply

We discuss each with a realistic example.

Apply

The ‘apply’ function is useful for producing results for a matrix, array, or data frame. They do this by producing results from the rows and or columns. The results of an ‘apply’ function are always shared as a vector, matrix, or list. Below is an example of the use of an ‘apply’ function.

You make a matrix that contains how many points James, Kevin, Williams have scored in the past three games. Below is the code for this.

> points.per.game<- matrix(c(25,23,32,20,18,24,12,15,16), ncol=3)
> colnames(points.per.game)<-c('James', 'Kevin', 'Williams')
> rownames(points.per.game)<-c('Game1', 'Game2', 'Game3')
> points.per.game
      James Kevin Williams
Game1    25    20       12
Game2    23    18       15
Game3    32    24       16

You want to know the most points James, Kevin, and Williams scored for any game. To do this, you use the ‘apply’ function as follows.

> apply(points.per.game, 2, max)
   James    Kevin Williams 
      32       24       16

Here is what we did

1. We used the ‘apply’ function and in the parentheses we  put the arguments “points.per.game” as this is the name of the matrix, ‘2’ which tells R to examine the matrix by column, and lastly we used the argument ‘max’ which tells are to find the maximum value in each column.
2. R prints out the results telling us the most points each player scored regardless of the game.

Sapply

The ‘apply’ function works for multidimensional objects such as matrices, arrays, and data frames. ‘Sapply’ is used for vectors, data frames, and list. ‘Sapply’ is more flexible in that it can be used for single dimension (vectors) and multidimensional (data frames) objects.  The output from using the ‘sapply’ function is always a vector, matrix, or list. Below is an example

Let’s say you make the following data frame

> GameInfo
  points      GameType
1    101          Home
2    115          Away
3     98 International
4     89          Away
5    104          Home

You now want to know what kind of variables you have in the ‘GameInfo’ data frame. You can calculate this one at a time or you can use the following script with the ‘sapply’ function

> sapply(GameInfo, class)
   points  GameType 
"numeric"  "factor"

Here is what we did

1. We use the ‘sappy’ function and included two arguments. The first is the name of the data frame “GameInfo” the second is the argument ‘class’ which tells R to determine what kind of variable is in the “GameInfo” data frame
2. The answer is then printed. The variable ‘points’ is a numeric variable while the variable ‘GameType’ is a factor.

lapply

The ‘lapply’ function works exactly like the ‘sapply’ function but always returns a list. See the example below

> lapply(GameInfo, class)
$points
[1] "numeric"

$GameType
[1] "character"

This the same information from the ‘sapply’ example we ran earlier.

Conclusion

The ‘apply’ family serves the purpose of running a loop without the concerns of the ‘for’ loop. This feature is useful for various objects.

In our last post on R, we looked at using the ‘switch’ function. The problem with the ‘switch’ function is that you have to input every data point manually. For example, if you have five different values you want to calculate in a function you need to input each value manually until you have calculated all five.

There is a way around this and it involves the use of a ‘for’ loop. A ‘for’ loop is used to repeat an action in R in a vector. This allows you to do a repeated action with functions that cannot use vectors.

We will return to are ‘CashDonate’  that has been used during this discussion of programming. Below is the code from R as it was last modified for the previous post.

CashDonate <- function(points, GameType){
 JamesDonate <- points * ifelse(points > 100, 30, 40)
 OwnerRate<- switch(GameType, Home = 1.5, Away = 1.3, International = 2)
 totalDonation<-JamesDonate * OwnerRate
 round(totalDonation)
}

The code above allows you to calculate how much money James and the owner will give based on points scored and whether it is a home, away, or international game. Remember, the problem with this code is that each value must be entered manually in the “CashDonate” function. To fix this we will used a ‘for’ loop as seen in the example below.

CashDonate <- function(points, GameType){
 JamesDonate <- points * ifelse(points > 100, 30, 40)
  OwnerRate <- numeric(0)
 for(i in GameType){
 OwnerRate<- c(OwnerRate, switch(i, GameType, Home = 1.5, Away = 1.3, International = 2))}
 totalDonation<-JamesDonate * OwnerRate
 round(totalDonation)
}

There are some slight yet significant changes to the function as explained below

1. The variable ‘OwnerRate’ is initially set to be a numeric variable with nothing inside it (0). This because different values will be put into this variable based on the ‘for’ loop.
2. Next, we have the ‘for’ loop. “i” represent whatever value is found in the argument ‘GameType’ the choices are home, away, and international. For every value in the dataframe the ‘CashDonate’ function will check the’ GameType’ to decide what to put into the ‘OwnerRate’ variable.
3. Next, we have the ‘OwnerRate’ variable again. We use the ‘c’ function to combine OwnerRate (which currently has nothing in it), with the “switch” function. The switch function looks at ‘i’ (from the ‘for’ loop) and calculates the same information as before.
4. The rest of the code is the same.

We will now try to use this modified ‘CashDonate’ function. But first we need a dataframe with several data points to run it. Input the follow dataframe into R.

GameInfo<- data.frame( points = c(101, 115, 98, 89, 104), GameType = c("Home", "Away", "International", "Away", "Home"))

In order to use the ‘CashDonate’ function with the ‘for’ loop you have to extract the values from the ‘GameInfo’ dataframe use $ sign. Below is the code with the answer from the ‘CashDonate’ function

CashDonate(GameInfo$points, GameInfo$GameType)
[1] 4545 4485 7840 4628 4680

You can check the values for yourself manually. The main advantage of the for loop is that we were able to calculate all five values at the same time rather than one at a time.

The ‘switch’ function in R is useful when you have more than two choices based on a condition. If you remember ‘if’ and ‘ifelse’ are useful when there are two choices for a condition. However, using if/else statements for multiple choices is possible but somewhat confusing. Thus the ‘switch’ function is available for simplicity.

One problem with the ‘switch’ function is that you can not use vectors as the input to the function. This means that you have to manually calculate the value for each data point yourself. There is a way around this but that is the topic of a future post. For now, we will look at how to use the ‘switch’ function.

Switch Function

The last time we discussed R Programming we had setup the “CashDonate” to calculate how much James would give for dollars per point and how much the owner would much James based on whether it was a home or away game. Below is the code.

CashDonate <- function(points, HomeGame=TRUE){
 JamesDonate<- points * ifelse(points > 100, 30, 40)
 totalDonation<- JamesDonate * ifelse(HomeGame, 1.5, 1.3)
 round(totalDonation)
}

Now the team will have several international games outside the country. For these games, the owner will double whatever James donates. We now have three choices in our code.

• Home game multiple by 1.5
• Away game multiple by 1.3
• International game multiple by 2

It is possible to use the ‘if/else’ function but it is complicated to code, especially for beginners. Instead, we will use the ‘switch’ function which allows R to switch between multiple options within the argument. Below is the new code.

CashDonate <- function(points, GameType){
 JamesDonate<- points * ifelse(points > 100, 30, 40)
 OwnerRate<- switch(GameType, Home = 1.5, Away = 1.3, International = 2)
 totalDonation<-JamesDonate * OwnerRate
 round(totalDonation)
}

Here is what the code does

1. The function ‘CashDonate’ has the arguments ‘points’ and ‘Gametype’
2. ‘JamesDonate’ is ‘points’ multiply by 30 if more than 100 points are scored or 40 if less than 100 points are scored
3. ‘OwnerRate’ is new as it uses the ‘switch’ function. If the ‘GameType’ is ‘Home’ the amount from ‘JamesDonate’ is multiplied by 1.5, ‘Away’ is multiplied by 1.3 and ‘International’ is multiplied by 2. The results of this is inputted into ‘totalDonation’
4. Lastly, the results in ‘totalDonation’ are rounded using the ’round’ function.

Since there are three choices for ‘GameType’ we use the switch function for this. You can input different values into the modified ‘CashDonate’ function. Below are several examples

> CashDonate(88, GameType="Away")
[1] 4576

> CashDonate(130, GameType="International")
[1] 7800

> CashDonate(102, GameType="Home")
[1] 4590

The next time we discuss R programming, I will explain how to use overcome the problem of inputting each value into the function manually.

In a previous post, we looked at If/Else statements in R. We developed a function that calculated the amount of money James and the owner would give based on how many points the team scored. Below is a copy of this function.

CashDonate <- function(points, Dollars_per_point=40, HomeGame=TRUE){
 game.points<- points * Dollars_per_point  if(points > 100) {game.points <-points * 30}
 if(HomeGame) {Total.Donation <- game.points * 1.5
 } else {Total.Donation <- game.points * 1.3}
 round(Total.Donation)
}

There is one small problem with this function. Currently, you have to input each game one at a time. You can have R calculate the results of several games at once. For example, look at the results of the code below when we try to input more than one game at once.

> CashDonate(c(99,100,78))
[1] 5940 6000 4680
Warning message:
In if (points > 100) { :
  the condition has length > 1 and only the first element will be used

As you can see, we get a warning message and some of the values are wrong. For example, the second value should be 4,500 and not 6,000.

In order to deal with this problem, R has the ‘ifelse’ function available. The ‘ifelse’ allows R to choose values in two or more vectors to complete an action. We need to be able to choose the appropriate action based on the following information

• points scored is less than or greater than 100
• Home game or not a home game

Remember, R could do this if one value was put into the ‘CashDonate’ function. Now we need to be able to calculate what to do based on several values in each of the vectors above. Below is the modified code for doing this.

CashDonate <- function(points, HomeGame=TRUE){
 JamesDonate<- points * ifelse(points > 100, 30, 40)
 totalDonation<- JamesDonate * ifelse(HomeGame, 1.5, 1.3)
 round(totalDonation)
}

Here is what the modified function does

1. It has the argument ‘points’ and the default argument of ‘HomeGame = TRUE’
2. The first calculation is the number of points but with the ‘ifelse’ function. If the number of points is greater than 100 the points is multiplied by 30 if less than 100 than the points are multiplied by 40. All this is put in the variable ‘JamesDonate’
3. Next, the amount from ‘JamesDonate’ is multiplied by 1.5 if it was a home game or 1.3 if it was not a home game. All this is put into the variable ‘totalDonation’
4. The results are rounded

To use CashDonate to its full potential you need to make a dataframe. Below is the code for the ‘games’ data frame we will use.

games<- data.frame(game.points=c(88,100,99,111,96), HomeGame=c(TRUE, FALSE, FALSE, TRUE, FALSE))

In the ‘games’ data frame we have two columns, one for game points and another that tells us if it was a home game or not. Now we will use the ‘games’ data frame with the new ‘CashDonate’ and calculate the results. We need to use the ‘with’ function to do this. This function will be explained at a later date. Below are the results.

> with(games, CashDonate(game.points, HomeGame=HomeGame))
[1] 5280 5200 5148 4995 4992

You can calculate this manually if you would like. Now, we can calculate more than one value in are ‘CashDonate’ function which makes it much more useful than before. All thanks to the use of the ‘ifelse’ function in the code.

Previously, we learned how to add nameless arguments to a function using ellipses ‘. . .’.  In this post, we will learn how to use functions as arguments in functions. An argument is the information found within parentheses ( ) or braces { } in R programming. The reason for doing this is that it allows for many shortcuts in coding. Instead of having to retype the formula for something you can pass the function as an argument in order to save a lot of time.

Below is the code that we have been working with for awhile before we add a function as an argument.

Percent_Divided <- function(x, divide = 2, ...) {
 ToPercent <- round(x/divide, ...)
 Output <- paste(ToPercent, "%", sep = "")
 return(Output)
 }

As a reminder, the ‘Percent_Divided’ function takes a number or variable ‘x’ divides it by two as a default and adds a ‘ % ‘ sign after number(s). With the ‘. . .’ you can pass other arguments such as ‘digits’ and specify how many number you want after the decimal.  Below is an example of the ‘Percent_Divided’ function in action for a variable called ‘B’ and with the add argument of ‘digits =3’

> B
[1] 23.35345 45.56456 32.12131
> Percent_Divided(B, digits = 3)
[1] "11.677%" "22.782%" "16.061%"

Functions as Arguments

We will now make a function that has a function for an argument. We will set a default function but remember that anything can be passed for the function argument. Below is the code for one way to do this.

Percent_Divided <- function(x, divide = 2,FUN_ARG = round, ...) {
 ToPercent <- FUN_ARG(x/divide, ...)
 Output <- paste(ToPercent, "%", sep = "")
 return(Output)
}

Here is an explanation

1. Most of this script is the same. The main difference is that we added the argument ‘FUN_ARG’ to the first line of the script. This is the place where we can insert whatever function we want. The default function is ’round’. If we do not specify any function ’round’ will be used.
2. In the second line of the code you again see ‘FUN_ARG’ this function will be activated after ‘x’ is divided by 2 and whatever arguments are used with the ‘. . .’
3. The rest of the code has already been explained and has not been changed
4. Important note. If we do not change the default of ‘FUN_ARG’, which is the ’round’ function, we will keep getting the same answers as always. The ‘FUN_ARG’ is only interesting if we do not use the default.

Below is an example of our modified function. The function we are going to pass through the ‘FUN_ARG’ argument is ‘signif’. ‘signif’ sounds the values in its first argument to the specified number of significant digits. We will also pass the argument ‘digits = 3’ through the ellipses ‘. . .’ The values of variable B (see above) will be used for the function

> Percent_Divided(B, FUN_ARG = signif, digits = 3)
[1] "11.7%" "22.8%" "16.1%"

Here is what happen

1. The Percent_Divided’ function was run
2. The function ‘signif’ was passed through the ‘FUN_ARG’ argument and the argument ‘digits = 3’ was passed through the ellipses ‘. . .’ argument.
3. All the values for B were transformed based on the script in the ‘Percent_Divided’ function

Conclusion

Using functions as argument is mostly for saving time when developing a code. Even though this seems complicated this is actually rudimentary programming.

In this post, we will continue the discussion on working with functions in R. Functions serve the purpose of programming R to execute several operations at once. This post, we will look at adding additional arguments to a function.

Arguments are the various entries within the parentheses. For example, in are example below the arguments of the function is x.

• MakePercent <- function(x) {
   ToPercent <- round(x, digits = 2)
   Output <- paste(ToPercent, "%", sep = "")
   return(Output)
  }

In the example above there are many other arguments beside x. However, the only argument for the function is x. The other arguments in other parentheses belong to  other objects in the script. In this post, we are going to learn how to add additional arguments to the function.

Let’s say that we want to convert a number to a percentage like a previous function we made but we now want to be able to divide the number by whatever we want. Here is how it could be done.

Percent_Divided <- function(x, divide) {
 ToPercent <- round(x/divide, digits = 2)
 Output <- paste(ToPercent, "%", sep = "")
 return(Output)
}

Here is what we did

1. We created the object ‘Percent_Divided’ and assigned the function with the arguments ‘x’ and ‘divide’
2. Next we use a { and we create the variable ‘ToPercent’ and we assigned the function ’round’  to round ‘x’ divided by whatever value ‘divide’ from the function takes. We then round the results of this two digits.
3. The results of ‘ToPercent’ are then assigned to the variable ‘Output’ where a ‘ %’ sign is assigned to the value
4. Lastly, the results of ‘Output’ are printed in the console.

Sounds simple. Below is the function in action dividing a number by 2 and then by 3

> source('~/.active-rstudio-document', echo=TRUE)

> Percent_Divided <- function(x, divide) {
+         ToPercent <- round(x/divide, digits = 2)
+         Output <- paste(ToPercent, "%", sep = "")
+    .... [TRUNCATED] 
> Percent_Divided(22.12234566, divide=2)
[1] "11.06%"
> Percent_Divided(22.12234566, divide=3)
[1] "7.37%"

Here is what happen

1. You source the script from the source editor by typing ctrl + shift + enter
2. Next, I used the function ‘Percent_Divided’ with the number 22.12234566 and I decided to divide the number by two
3. R returns the answer 11.06%
4. Next I repeat the process but I divide by 3 this time
5. R returns the answer 7.37%

There is one problem. The argument ‘divide’ has no default  value. What this means is that you have to tell R what the value of ‘divide’ is every single time. As an example see below

> Percent_Divided(22.12234566)
Error in Percent_Divided(22.12234566) : 
  argument "divide" is missing, with no default

Because I did not tell R what value ‘divide’ would be, R was not able to complete the process of the function. To solve this problem we will set the default value of ‘divide’ to 10 in the script as shown below.

Percent_Divided <- function(x, divide = 10) {
 ToPercent <- round(x/divide, digits = 2)
 Output <- paste(ToPercent, "%", sep = "")
 return(Output)
}

If you look closely you will see ‘divide = 10’. This is the default value for ‘divide’ if we do not set another number for ‘divide’ R will use 10. Below is an example using the default value of ‘divide’ and another example with ‘divide’ set to 5.

> Percent_Divided <- function(x, divide = 10) {
+         ToPercent <- round(x/divide, digits = 2)
+         Output <- paste(ToPercent, "%", sep = "") .... [TRUNCATED] 
> Percent_Divided(22.12234566)
[1] "2.21%"
> Percent_Divided(22.12234566, divide = 5)
[1] "4.42%"

First we sourced the script using ctrl + shift + enter. In the first example, the number is automatically divided by 10 because this is the default. In the second example, we specific we wanted to divide by five by adding the argument ‘divide = 5’. You can see the difference in the results.

In a future post, we will continue to examine the role of arguments in functions.

In this post, we will explore how to create data frames as well as looking at other aspects of using data frames in R. The first example below is a data frame that contains information about fictional faculty members. Our job will be to put this information a data frame and to rename the columns. Below is the example and it will be followed by an explanation.

> Faculty <- c('Darrin Thomas', 'Hank Smith', 'Sarah William')
> Salary <- c(60000, 50000, 53000)
> Hire_Date <- as.Date(c('2015-1-1', '2000-6-1', '2012-9-1'))
> Lecturers.data <- data.frame(Faculty, Salary, Hire_Date)
> str(Lecturers.data)
'data.frame':	3 obs. of  3 variables:
 $ Faculty  : Factor w/ 3 levels "Darrin Thomas",..: 1 2 3
 $ Salary   : num  60000 50000 53000
 $ Hire_Date: Date, format: "2015-01-01" "2000-06-01" "2012-09-01"

Here is what happen

1. We started by making three different vectors and assigning a variable to each. The variables are ‘Faculty’, ‘Salary’, and ‘Hire_Date’.
2. We then assigned all three variables to the data frame ‘Lecturers.data’ ‘Faculty’ is a factor vector, ‘Salary’ is a numeric vector, and ‘Hire_Date’ is a date vector. Again, the advantage of data frames is their ability to have several different types of data
3. We then used the ‘str’ function to see the attributes of the ‘Lecturers.data’ data frame.

There is one small problem with the data frame above. ‘Faculty’ is a factor vector but our original vector for “Faculty’ was a character vector. We want ‘Faculty’ to continue to be a character vector instead of it becoming a factor. The example below shows one way to deal with this small problem.

> Lecturers.data <- data.frame(Faculty, Salary, Hire_Date, stringsAsFactors=FALSE)
> str(Lecturers.data)
'data.frame':	3 obs. of  3 variables:
 $ Faculty  : chr  "Darrin Thomas" "Hank Smith" "Sarah William"
 $ Salary   : num  60000 50000 53000
 $ Hire_Date: Date, format: "2015-01-01" "2000-06-01" "2012-09-01"

By adding the argument ‘stringsAsFactors=FALSE’ it make forces all vectors to not be factors. If you look closely you will see that ‘$ Faculty’ is not a Factor anymore as in the previous example. Instead it is now a ‘chr’ or character variable.

It is also possible to rename column names in a data frame just like in a matrix. For example, let’s say you made a mistake with the ‘Hire_Date’ variable. You did not mean the the date the lecturers were hired but the date they resigned. Below is an example of how to fix this.

> Lecturers.data
        Faculty Salary  Hire_Date
1 Darrin Thomas  60000 2015-01-01
2    Hank Smith  50000 2000-06-01
3 Sarah William  53000 2012-09-01
> names(Lecturers.data) [3] <- 'Resign_Date'
> Lecturers.data
        Faculty Salary Resign_Date
1 Darrin Thomas  60000  2015-01-01
2    Hank Smith  50000  2000-06-01
3 Sarah William  53000  2012-09-01

Here is what happening

1. We displayed the data frame ‘Lecturers.data’ as a reference point
2. We noticed that we did not want a column named ‘Hired_Date’ but want to change the name to ‘Resign_Date’
3. To change the name we use the ‘names’ function to change the name of a column in ‘Lecturers.data’. We specifically tell are to rename the third column by using the subset brackets [3] and assign the name ‘Resign_Date’
4. We then redisplay the ‘Lecturers.data’ data frame. If you compare this data frame with the first you can see that the third column has been renamed as desired.

This post provided some basic information on developing data frames. We learned how to combine vectors into a data frame, how to change a factor to a character vector, and how to rename a column. Such skills as these are beneficial to anyone who needs to use data frames.

So far we have looked at vectors, matrices, and arrays. One thing these three objects have in common is that they consist of one type of data. In other words, vectors, matrices and arrays contain either numerical or character information but not both at the same time.

Data frames are different. They allow you to have a mixture of information all contained within one place. You can have character data, such as names, and numerical data, such as salaries all in one place. In this post, we will first look at how to convert a matrix to a data frame.

Converting a Matrix to a Data Frame

One of the benefits of converting a matrix to a data frame is that the columns in a matrix become variables in a data frame. This is useful for data analysis at times. In the example below, we will convert the matrix of ‘points.team’ into a data frame.  In order to do this we have to use to new functions the ‘as.data.frame’ function which converts the matrix into a data frame and the ‘t’ function which transposes the rows so that they become the columns. Below is the code for completing this

> points.team
      1st 2nd 3rd 4th 5th 6th
James  12  15  30  25  23  32
Kevin  20  19  25  30  31  22
> team.points.df <- as.data.frame(t(points.team))
> team.points.df
    James Kevin
1st    12    20
2nd    15    19
3rd    30    25
4th    25    30
5th    23    31
6th    32    22

This is what we did

1. We displayed the matrix ‘points.team’ as a reference. The code for creating this is available here.
2. We then created the variable ‘team.points.df’ and used two functions to convert the matrix ‘points.team’ to a data frame
   1. ‘as.data.frame’ was used to make the actually data frame
   2. ‘t’ was used to move or transpose the names of the rows in the matrix (James and Kevin) to be the names of columns in the data frame. This gives us two variables (James and Kevin) with six entries (1st-6th) if we had not done this we would have made the example below

      > team.points.df
            1st 2nd 3rd 4th 5th 6th
      James  12  15  30  25  23  32
      Kevin  20  19  25  30  31  22

      In this example, we did not transpose ‘James’ and ‘Kevin’ to be columns. Instead 1st-6th are the variables instead of ‘James’ and ‘Kevin’. For our purposes this does not make sense but it may be appropriate it other situations.

3. The last step involved displaying the new data frame ‘team.points.df’

Using the ‘str’ function allows you to learn some information about a data frame as in the example below

> str(team.points.df)
'data.frame':	6 obs. of  2 variables:
 $ James: num  12 15 30 25 23 32
 $ Kevin: num  20 19 25 30 31 22

Here is what we now know about our data frame

• The variable is a data frame (data.frame)
• It has six observations (6 obs.) which are the points scored by the players in six games
• There are two variables ‘James’ and ‘Kevin’
• The variables are numeric (num)

This is just the beginning of our examination of data frames in R. In a future post we will look at making original data frames.

In this post, we will look at how to rename the rows and columns in a matrix. We will also examined how to do the following

• Name rows and columns in matrices
• Make an array
• Basic math in matrices & arrays

Naming Rows and Columns

Renaming rows and columns has a practical use. It allows people to provide meaning and or context to the data they are interpreting. In the example below, we will rename the rows and the columns of the basketball players so that it contains their names as well as what game.

> points.of.James
[1] 12 15 30 25 23 32
> points.of.Kevin
[1] 20 19 25 30 31 22
> points.team <- rbind(points.of.James, points.of.Kevin)
> points.team
                [,1] [,2] [,3] [,4] [,5] [,6]
points.of.James   12   15   30   25   23   32
points.of.Kevin   20   19   25   30   31   22
> rownames(points.team) <- c("James", "Kevin")
> points.team
      [,1] [,2] [,3] [,4] [,5] [,6]
James   12   15   30   25   23   32
Kevin   20   19   25   30   31   22
> colnames(points.team) <- c("1st", "2nd", "3rd", "4th", "5th", "6th")
> points.team
      1st 2nd 3rd 4th 5th 6th
James  12  15  30  25  23  32
Kevin  20  19  25  30  31  22

.Here is what’s going on

1. We make the variables ‘point.of.James’ and ‘point.of.Kevin’ and display them
2. We combine ‘point.of.James’ and ‘point.of.Kevin’ into a matrix using the ‘rbind’ function and assign this to the variable ‘points.team’
3. We then display ‘points.team’
4. We then rename the rows using the ‘rownames’ function. We replace ‘point.of.James’ and ‘point.of.Kevin’ with ‘James’ and “Kevin’ in the rows. We then display this change in the matrix.
5. Next we rename 1,2,3,4,5,6 with 1st, 2nd, 3rd, 4th, 5th, & 6th. using the ‘colnames’ function.
6. Lastly, we display the finished table.

Making Arrays

A vector has one dimension (row), a matrix has two dimensions (row & column), and an array has three or more dimensions (length, width, & height for example). We are not going to cover arrays extensively because it is hard to envision data beyond 3 dimensions. In addition, please remember that all the rules and tricks for vectors and matrices apply towards an array. Below is an example of how to make an array

> array1 <-array(1:48, dim=c(6, 8, 4))
> array1
, , 1

     [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8]
[1,]    1    7   13   19   25   31   37   43
[2,]    2    8   14   20   26   32   38   44
[3,]    3    9   15   21   27   33   39   45
[4,]    4   10   16   22   28   34   40   46
[5,]    5   11   17   23   29   35   41   47
[6,]    6   12   18   24   30   36   42   48

, , 2

     [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8]
[1,]    1    7   13   19   25   31   37   43
[2,]    2    8   14   20   26   32   38   44
[3,]    3    9   15   21   27   33   39   45
[4,]    4   10   16   22   28   34   40   46
[5,]    5   11   17   23   29   35   41   47
[6,]    6   12   18   24   30   36   42   48

, , 3

     [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8]
[1,]    1    7   13   19   25   31   37   43
[2,]    2    8   14   20   26   32   38   44
[3,]    3    9   15   21   27   33   39   45
[4,]    4   10   16   22   28   34   40   46
[5,]    5   11   17   23   29   35   41   47
[6,]    6   12   18   24   30   36   42   48

, , 4

     [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8]
[1,]    1    7   13   19   25   31   37   43
[2,]    2    8   14   20   26   32   38   44
[3,]    3    9   15   21   27   33   39   45
[4,]    4   10   16   22   28   34   40   46
[5,]    5   11   17   23   29   35   41   47
[6,]    6   12   18   24   30   36   42   48

Here is what is happening

1. We created ‘array1’ and assigned an array to it that contains numbers 1 to 48 and has 6 rows, 8 columns, and 4 dimensions.
2. When then have R display the results

To locate values in specific indices you have to add an additional number to the address. For example, if you are looking for 1,1,1 you would go to the first dimension, first row, and first column. Also do not forget how extracting rows and columns re-shuffles the numbering of the remaining rows and columns.

Basic Math in Matrices and Arrays

You can change all the values in a matrix and array. For example, let us say we want to add four points to every game that Kevin and James played. We can do this by using the following code.

> points.team
      1st 2nd 3rd 4th 5th 6th
James  12  15  30  25  23  32
Kevin  20  19  25  30  31  22
> new.points.tream <- points.team+4
> new.points.tream
      1st 2nd 3rd 4th 5th 6th
James  16  19  34  29  27  36
Kevin  24  23  29  34  35  26

What happening

1. We redisplayed the variable “points. team’
2. We then create a new variable called ‘new.points.team’ and we told R to add 4 points to every value in the variable ‘points.team’
3. We display the new table. A closer look will show how every value is 4 points higher in the new table than the results of

Conclusions

This conclude the examinations on matrices and arrays.

In this post, we will continue to explore the basic features of matrices. Specifically, we will look at how to replace values in a matrix and how to combine vectors into a matrix.

Replacing Values in a Matrix

There are times when we might input the wrong value into a matrix. The hard way to deal with this problem is to remake the entire matrix. The easy way is to only replace the incorrect value. You can change values by individual index, by row or column or by even importing another matrix (we will not covering replaces values in a matrix with another matrix). Below are examples

> matrix1<-matrix(1:20, ncol=4)
> matrix1
     [,1] [,2] [,3] [,4]
[1,]    1    6   11   16
[2,]    2    7   12   17
[3,]    3    8   13   18
[4,]    4    9   14   19
[5,]    5   10   15   20
> matrix1[1,1] <- 4
> matrix1
     [,1] [,2] [,3] [,4]
[1,]    4    6   11   16
[2,]    2    7   12   17
[3,]    3    8   13   18
[4,]    4    9   14   19
[5,]    5   10   15   20

In this first example we wanted to replace the value in index 1,1 with the number 4 here is what happened.

1. We created the matrix ‘matrix1’ using numbers 1-20 (1:20) with 4 columns (ncol = 4).
2. We type ‘matrix1’ so R displays it. The purpose of this is so that we can compare the original matrix with the modification.
3. We than type in ‘matrix1’ into R and subset row 1 column 1 using brackets and we assigned the number 4 to row 1 column 1.
4. We then type ‘matrix1’ to show the modified matrix

If you look carefully at 1,1 in the matrix you will that the value in this index has been changed from 1 to 4.

Below is an example of replace an entire row. This technique can also be used for replacing a column

> matrix1<-matrix(1:20, ncol=4)
> matrix1
     [,1] [,2] [,3] [,4]
[1,]    1    6   11   16
[2,]    2    7   12   17
[3,]    3    8   13   18
[4,]    4    9   14   19
[5,]    5   10   15   20
> matrix1[2, ] <- c(2,3)
> matrix1
     [,1] [,2] [,3] [,4]
[1,]    1    6   11   16
[2,]    2    3    2    3
[3,]    3    8   13   18
[4,]    4    9   14   19
[5,]    5   10   15   20

Here is what happened

1. We created the matrix ‘matrix1’ and displayed it as a reference
2. We then subsetted the second row of ‘matrix1’ and told are to put in that row the numbers 2 and 3
3. We then displayed the modified matrix

If you look carefully, you will see that in the second row we now have the pattern 2,3.  When you tell R to replaces values using a pattern, R will continue to repeat the pattern until the row is filled. Even though we only told R to use two numbers (2 and 3) R filled all four columns in the second row with the pattern 2,3.

Combining Vectors into a Matrix

You can combine two or more vectors into a matrix by using the function ‘rbind’ below is an example.

> points.of.James
[1] 12 15 30 25 23 32
> points.of.Kevin
[1] 20 19 25 30 31 22
> points.of.players <- rbind(points.of.James, points.of.Kevin)
> points.of.players
                [,1] [,2] [,3] [,4] [,5] [,6]
points.of.James   12   15   30   25   23   32
points.of.Kevin   20   19   25   30   31   22

Here is what we did

1. We redisplayed the values of the variables ‘points.of.James’ and ‘points.of.Kevin’
2. When then created the variable ‘points.of.players’ and used the ‘rbind’ function to combine the values in the vectors/variables of ‘points.of.James’ and ‘points.of.Kevin’
3. We then display the results

In a future post, we will examine how to rename the rows and columns so that they provide critical information for people who may be trying to interpret the information.

In this post, we will continue to explore the different actions that can be performed when using matrices and arrays. In particular, we will look at how to manipulate the values within a matrix.

Manipulating Values in a Matrix

Before manipulating values it is important to understand what an index is. An index or Indices, which is the plural form, is the address f a value within a matrix. There are two numbers involved in an index. The first number represents the row that the value is in and the second number represents the column. Again this is similar to Microsoft excel and its system for giving every cell an address.

Below is an example to help make this clearer

> matrix1<- matrix(1:20, ncol=4)
> matrix1
     [,1] [,2] [,3] [,4]
[1,]    1    6   11   16
[2,]    2    7   12   17
[3,]    3    8   13   18
[4,]    4    9   14   19
[5,]    5   10   15   20

Here is what we did,

1. We made a variable called ‘matrix1’ and this variable has a matrix of 20 consecutive numbers (1:20) with four columns (ncol = 4)
2. We then print ‘matrix1’ in R

Now for a simple quiz

What value is in row 3 column 2?*

What value is in row 5 column 4?**

The answers are at the bottom of the post.

In R, we do not say ‘row 3’ and ‘column 2’. Instead we would simply say 3, 2 or 5, 4 in the code.

Extracting Values

With our knowledge of indices we can now extract or removes values from a matrix. For example, below is a way to extract and only look at the first two rows and columns two and three of our matrix. In each example, I will display the original matrix and then the extracted matrix so you can compare the differences.

> matrix1
     [,1] [,2] [,3] [,4]
[1,]    1    6   11   16
[2,]    2    7   12   17
[3,]    3    8   13   18
[4,]    4    9   14   19
[5,]    5   10   15   20
> matrix1[1:2, 2:3]
     [,1] [,2]
[1,]    6   11
[2,]    7   12

Here is what happen

1. We printed matrix1 as a reference for comparison
2. We then extracted the first two rows (1:2) and the second and third columns (2:3) of ‘matrix1’ using brackets
3. R then prints the results

Notice how R renames the rows and columns. Rows 1 and 2 are still rows 1 and 2 because we extracted them. However, what use to be rows 2 and 3 has been renamed rows 1 and 2 in the extracted matrix. This can get really confusing so be careful.

If you want to extract an entire row you do not specify the column as in the example below

> matrix1
     [,1] [,2] [,3] [,4]
[1,]    1    6   11   16
[2,]    2    7   12   17
[3,]    3    8   13   18
[4,]    4    9   14   19
[5,]    5   10   15   20
> matrix1[2:3,]
     [,1] [,2] [,3] [,4]
[1,]    2    7   12   17
[2,]    3    8   13   18

Here is what happening.

1. We print ‘matrix1’ as a reference for comparison
2. We extract from ‘matrix1’ rows 2 and 3 (2:3) using brackets. Notice that we put a comma after 2:3. This tells R to take the whole row.
3. R prints the results. Notice what was once rows 2 and 3 is now rows 1 and 2.

You can also remove rows and columns in a matrix as in the example that follows

> matrix1
     [,1] [,2] [,3] [,4]
[1,]    1    6   11   16
[2,]    2    7   12   17
[3,]    3    8   13   18
[4,]    4    9   14   19
[5,]    5   10   15   20
> matrix1[-2,-3]
     [,1] [,2] [,3]
[1,]    1    6   16
[2,]    3    8   18
[3,]    4    9   19
[4,]    5   10   20

Here is what is going on

1. We print ‘matrix1’ as a reference.
2. We then tell R to remove row 2 and column 3 (-2, -3). The removal of a row and or column is indicated through using a negative – sign.
3. R prints the results. Notice that there are now 4 rows and 3 columns because we remove one row and one column. In addition, don’t forget that the numbers and columns have been renumbered

ANSWERS TO QUIZ

*8

**20

Vectors, matrices, arrays, what is the difference? The difference has to do with the number of dimensions that they have. A vector is only only one dimension. Vectors are one row of information. Matrices have two dimensions which are rows and columns. Any Microsoft excel spreadsheet is a matrix of rows and columns and thus contains two dimensions. An array is anything with three or more dimensions such as height, width and depth. Anything beyond three dimensions is hard for the typical mind to grasp.

Creating a Matrix

We will now create a simply matrix below. An explanation is provided after the code

> matrix.1 <- matrix(1:20, ncol=4)
> matrix.1
     [,1] [,2] [,3] [,4]
[1,]    1    6   11   16
[2,]    2    7   12   17
[3,]    3    8   13   18
[4,]    4    9   14   19
[5,]    5   10   15   20

Here is what we did

1. We created the variable ‘matrix.1’
2. Within this variable we put a matrix that contain the numbers 1 to 20 (1:20)
3. When the specified that we want four columns.
4. R creates the matrix
5. We type ‘matrix.1’ into R and press enter so we can see the matrix

Once you specify the number of columns R determines the number of rows itself. However, R does not like empty space in the rows and columns. Therefore, you matrix must use all of the space possible within it or you will get the message below.

> matrix.1 <- matrix(1:19, ncol=4)
Warning message:
In matrix(1:19, ncol = 4) :
  data length [19] is not a sub-multiple or multiple of the number of rows [5]

In the example above, I tried to put 19 numbers into a matrix that had 4 columns and 5 rows which equals 20 spaces or indices. Since there was one empty space R did not want to create the matrix.

Examining the Properties of Matrices

If you type in the function “str” you get the following information

> str(matrix.1)
 int [1:5, 1:4] 1 2 3 4 5 6 7 8 9 10 ...

Here is what this information means

• int means that the matrix contains integers
• 1:5 means that there are 5 rows and all rows are included in this description
• 1:4 means that there are 4 columns and all columns are included in this description

• 1 2 3 4 5 6 7 8 9 10 … is just the first ten indices in the matrix

The “dim” function will tell you the dimensions of the matrix and the “length” function tells you how many indices there are in the matrix. Both are below.

> dim(matrix.1)
[1] 5 4
> length(matrix.1)
[1] 20

The “dim” function indicates we have 5 rows and 4 columns. The ‘length’ function indicates that we have 20 indices.

This serves only as in introduction to matrices. We haven’t really dealt with arrays yet but many of the same ideas and concepts apply equally to them as well.