1. Bar charts
In this next section you will work with bar charts. These simple visualizations can be an effective tool for comparing frequencies of data including factors like eye color "brown", "blue", "hazel" and so on.
2. How do you visualize non-numerics?
Sometimes you will have non-numeric data that you want to explore visually. The most common way to do this is by counting each type of non-numeric level. This data shows eye colors. You can't equate 1 to Blue and 2 to Brown eyes. But you can count each eye color and build this "bar chart" which sums up the number of blue or brown eyes in your variable. Simply highlight the single column then insert a bar chart in the dialog.
3. Geeked Up Off Them Bars!!
If you have numeric data, or can summarize levels through a pivot table you can also use bar charts. The data now has gender along with eye color. Now when you insert a bar chart you can compare gender by each category side by side. Use a side by side bar chart when you want to compare quantities by category like Brown Eyes.
4. Still another way
You can modify your bar chart from side by side to a stacked bar chart shown here. A stacked bar chart lets your brain identify the proportion of a variable within a level as part of the total. In contrast, side by side bar charts lets you compare quantities not proportionally among the category level. This is the same gender and eye color data visualized in a stacked manner. Now you can see that proportionally Brown eyes are about even because the blue and red have similar areas totally almost 150.
5. Proportional Stacked charts
Lastly, another way to review this data is with a proportional stacked bar chart. Instead of raw frequencies, each column is colored by its proportion of the total in that category level. This view allows your audience to comprehend the proportion split in comparison to other categories even if the quantities themselves are small. In previous visuals the "other" bars were smaller because the values were small in comparison to the other categories. This made it hard to compare in a side by side or a stacked bar chart. However in a proportional stacked bar chart, all bars are the same height. This let's your brain focus on making categorical comparisons from brown, to blue to other not just within a single category. Within the *other* category now you can see that the red bar is longer than blue. Even though there wasn't many observations in the *other* category this view shows a difference more clearly than previously.
6. Categories in an analysis
In this chapter, data is related to auctions and their such as seller ratings & closing price. In this auction data, there is a `Category` column with no numeric information. Instead these are attributes of the auctions. In fact they are distinct classes, known as factors. You often learn a lot about the descriptive statistics of your data when you understand the frequency of meaningful factor and how it relates to another variable. So exploring a category and its relationship to other variables can be informative. For example you may explore your data to find out that sporting good auctions have a higher proportion of competitive auctions compared to other auction types. In fact, making multiple bar charts as part of your real world auction analysis will help you spot these patterns in the data.
7. Let's practice!
You're doing great. Keep up the good work.