Excel is a very popular tool to create graphs. In addition, ASP comes with a set of graphical tools for data graphs. In this video, we briefly discuss graphs in Excel and ASP. We also take this opportunity to discuss common ways to encode numerical values and compare them. In the second course of this specialization we discussed six graphs for exploratory data analysis. For categorical data, we introduced the bar graphs and pie charts. For quantitative data, we introduced histograms, boxplots and timeplots. In addition, we also introduced scatter plot as a way to examine relationships between a pair of quantitative variables. Excel does not include queues for all these graphs, in particular histograms and boxplots cannot be easily created in Excel, even though it is possible to create them by following some carefully planned steps. The graphing tools in ASP can create all these graphs. Excel comes with many other graphs that were not mentioned before, such as tree map, pivot chart, waterfall chart, and the win/loss graph. We choose not to discuss them in detail here. Several additional graphs are available in ASP, including histograms and boxplots I mentioned earlier. Even though, I choose not to go into details, I suggest that you explore them. If you are reasonably comfortable with creating graphs in Excel, creating graphs in ASP should be easy to master. An important decision with creating graphs in Excel is the way to encode data. There are many alternatives including points, lines, bars, boxes, shapes, and color intensity. Points are also use to represent a single value as in a dot plot or pair of values as in a scatter plot. Lines are primarily used for two purposes. First, it can be used to collect individual values, second, it can also be used to display overall pattern. An important example is the trend line in a scatter plot. Bars are typically used to represent a single value. Note that even though bars are two dimensional objects with both height and width, one of the dimensions, the width is typically not used to encode a data value. Boxes are often used to represent ranges of values. The most well known example is a boxplot. It is also possible to use shapes represent single values. For example, a larger circle represent a bigger value. Finally, it is also possible to use color intensity to represent a single values. Usually, darker colors are used to represent higher value. And the last two, shapes and color intensity are four last common in the first four. Now, let me discuss two examples to illustrate the concentrations when choosing the way to encode data. The first example concerns time series data. Here, I show the same time series data using three different ways to enclose the data values. Points are using the first one. Bars are using the second one and the lines are using the last one. Which one is better? Experts often suggest that we avoid using points. Well, bar and lines can both be appropriate they serve different purposes. Bars are often used when we want to emphasize individual values and the lines are used when we want to emphasize a trend. Here's the rationale for the recommendation. Points have small visual weight and do not help show the sequential nature of the time series data. Bars have large visual weight and make individual values stand out. Finally, adding a line helps correct the issue. Therefore, depending on the purpose of our graph we should carefully choose the type of graph we use. In Excel, if we want to use bars we need to choose bar graph. If we want to use lines we need to choose line graph. Our second example concerns graphs used to show distribution of data. Recall that a distribution shows how a set of data spreads across the entire range. You can choose to use several different ways to encode the data. If we use points the graph is typically called strip plot. If we use lines, the graph is called density curve. If we use bars, the graph is called histogram. Finally, if we use boxes, the graph is called boxplot. Note that the different ways to show distribution have different strengths. The strip plot uses points. It is great for showing individual values in the data set. However, it becomes infeasible when data set is relatively large. The line in the density curve features the shape of the distribution. The bars in the histogram naturally shows the frequency of each interval within the data range. Boxplot is an exceptionally simple graph that features simple summary statistics of the data. Which one we should choose depends on the purpose. Therefore, histograms and the boxplots are most popular. Histograms show a lot more details than boxplots. However, boxplots can be great choice as it succinctly summarizes a data set and are also great for side-by-side comparison for several related data sets.