TDM 10100: Project 4 - Data Types

Vectors are fundamental data structures and are used frequently for storing and manipulating data. A vector contains an ordered collection of values, all of the same type of data.

A numeric vector would contain only numerical values
my_num <- c(1, 2, 3, 4, 5, 6, 7)
Running my_num outputs: 1 2 3 4 5 6 7

A character vector would usually contain only string values
my_str <- c("This", "is", "my", "vector")
Running my_str outputs: 'This' 'is' 'a' 'string'

Combining the numeric vector my_num and the character vector my_str does not work well
c(my_num, my_str)
Output: '1' '2' '3' '4' '5' '6' '7' 'This' 'is' 'a' 'string'

The numeric data gets converted to strings to make all of the values in the vector of one type.

Read in the titanic dataset using one of the following three methods: Download the file from the GitHub titanic.csv file found here, upload it to your Jupyterlab files, and read it into your notebook using read.csv and your personal file path.

With the titanic data read in as myDF, look at the str() (structure) and summary() of the dataset. This shows the type of data within each column.

Create two vectors: my_names and my_ages, containing the Name and Age columns, repectively, of myDF.

Combine my_names and my_ages to create a new vector.

An alternative to creating a vector where the data has been converted to be of one type, you could also create a two-column dataframe containing just the columns you have selected. This isn’t a vector and would serve a different purpose, but is just also another way to pull some data and create a subset.

Look at the dimensions of this new dataframe to confirm you have 891 rows, 2 columns.

It is easy to learn a lot about data just by looking at how it is made up. Checkout the length, class, type of, and entry from the 45th row of both my_names and my_ages.

my_ages is of type "double". Data can be of the class numeric. When this is the case, it is stored as either a "double" (decimal) or "integer" (whole number) internally.

Looking at the specific row entry with my_ages[45], we’re pulling a single row value. R prints it in the format most simple for us to read 19 instead of 19.00, because this is a whole number and does not need the decimal values. But this is still stored as a "double". Why?

If you use print() to display the values of my_ages, even whole numbers like 19 will be displayed in decimal.
When we look at more entries from this column, such as all ages listed as less than 20 years old, there are entries such as 14.5 that show there are, indeed, decimal values in this column. This makes all the data in this column be of that type, but for good reason because these values weren’t rounded just to be stored a certain way.

In a similar format, we can find the names that go along with each of the "under 20" ages.

Save the non-NA values of my_ages that are under 20 as my_selection.

Use paste() to bring together my_ages and my_names filtered by my_selection. Save this as age_names and print().

Looking at the Survived column of the titanic dataset, we can see that it contains binary values for each person’s life status:

We’re going to create a column containing the life status and sex of each person. This allows for future data analysis when looking at the counts of what sort of people survived, and so on.

But first it would be helpful to convert these values from 0/1 to Dead/Alive. This column is numeric data, and we want to make a new column containing labels for each value to make it easier to understand.

The factor() function takes the original vector - often numeric or character (the Survived column in this case) - splits up based on the unique values, and applies a label to each.

For example, if we wanted to split the Pclass (passenger class) column and add labels, we could factor() the Pclass based on each of the three choices (1, 2, 3), relabel them (First Class, Second Class, Third Class), and save this as a new column. Using the basic structure:

Make sure to fill out this code with the class values and the label you want each to have.

The class() of Passenger_Class is "factor", but the typeof() remains "integer" as with Pclass. This column still contains representation for those 1, 2, 3 values, just with the class labels for each numerical value.

Returning to the Survived column, use factor() to create a new column Status containing "Dead" and "Alive" labels on the values.

Now we get to combine this new Status column with the Sex column to create Combined. paste() makes this easy and fairly painless. It is totally up to you for what you would like to have as the separator between the two values in each row. Some commonly used ones are:

Anything could be used, but these are what you would commonly see, and are often used to separate words.

Make sure to view some of this column to ensure everything looks how you would like it to. When you look at the table of this Combined column, check that it contains all four possible combinations.

Often it is helpful (or even fun) to make a visual to go along with findings from a table. Please use a barplot() to show the values from this table visually, and customize as you would like. There should be a bar for each category:

You can sort the data, add different colors, rotate the plot, rotate the labels, etc. Just make sure there are axis labels and a title that make sense to what you are showing.

The end goal of these last two questions is to create a plot that could provide some insight to how the age and sex of a person relate to whether or not they survived, and how common each of these occurrences are.

We’re going to be working with the Age column now. There are the ages of the passengers that range from 0.42 to 80 (make sure to find this yourself!), with many values in between.

Vectors are often shown in examples where they are taking numbers like 1, 2, 3, 4 and combining them to show how a vector can be created. These don’t really allow for understanding of how or why this would be done in the real-world, especially when working with big datasets.

With the titanic data in myDF, we’re going to look at the table of the Age column and choose a cut-off for what counts as "old". This will include the ages 61 - 80.

Create a vector old_ages containing the ages 61, 62, 63, 64, 65, 66, 70, 70.5, 71, 74, and 80.

The values in the old_ages vector match those in the upper values of the Age column, but are not directly tied to the dataset just yet. Printing this vector just shows the numbers 61, 62, 63, 64, 65, 66, 70, 70.5, 71, 74, and 80, with no counts related to any of them.

As fun as it was to have a long section of text saying 61, 62, 63, 64, 65, 66, 70, 70.5, 71, 74, and 80, it is often simpler to use a range when selecting values like this.

What we’re about to do is completely non-efficient but it gives us some practice working with and manipulating vectors.

First, create a vector my_vec1 containing all the values from the Age column that are less than or equal to 10 and print. Put that aside, and create a new vector my_vec2 containing a range of all the values from the Age column that are greater than 10 and go to 60.

Remember to remove NA values!

With my_vec1 and my_vec2, combine them to make the vector people.

From people, we need to create three more vectors:

To add labels to the Age column that correspond to these different ranges, use %in% again to filter each entry of the Age column and assign each value a name. Save this as AgeGroup rather than overwriting the Age column.

It is often the case that it is useful to create a visualization to help better understand comparisons. In the AgeGroup column, we have different counts of occurrences of each age range from the Age column.

In the Combined column, there are different counts of each survived Status (dead or alive) sorted by Sex. Having these values is great, but we haven’t done any analysis with them yet.

Going back to that plot we wanted to make since the start of Question 4, we are going to use AgeGroup and Combined to make it.

To start, make one table showing both the AgeGroup and Combined together. You usually make a table showing the counts for each unique value in one column. Here, it is useful to show the unique values from one column, and then those from another, and cross them to get the counts from each pairing.

This will give us some insight about what sorts of people survived (or didn’t). This is not a huge dataset, so the values can be on the smaller side, but this does make it easier to fully grasp what the changes in values are.

Save the table as a new variable, and plot it.

There are some choices when it comes to this:

You are welcome to use any or all plotting methods (make two plots), and can bring in your own as you see fit. Make sure to include a legend when the color values are important to reading the plot. Also, the values from Combined are rather long, so you may have to adjust the margins values some.