In this demo, I'll use the automated explanation in Visual Analytics to understand the underlying factors for Threat Level. I'm already signed in and I have the VA2-Demo1.2 report open in SAS Visual Analytics. I want to start by viewing the values for Threat Level. In the left pane, I'll click the Data icon, find Threat Level in the Category group, right-click the data item and select Edit. Threat Level is a custom category based on conservation status. On the left side of the window, I can see two values for conservation status: missing and Species of Concern. A conservation status of (missing) indicates that the species has not been identified as not threatened, threatened, or extinct. A conservation status of Species of Concern indicates that the species might need proactive protection, but there is insufficient information available to list the species as endangered. On the right side of the window, I see other values of conservation status that have been grouped into not threatened, threatened, and extinct, respectively. Any values that are not grouped on the right side of the window will be displayed as is: with their original values. When I'm done looking at the custom category, I'll cancel out of the window. Now, I would like to use the automated explanation to explain the characteristics and contributing factors for threatened species. In the Data pane, I'll right-click Threat Level and select Explain > Explain on current page. When the automated explanation is created, Threat Level is specified as the response level, and most of the remaining data items in the table are automatically added as underlying factors. Before I view specifics about which factors are included and which have been rejected during variable screening, I want to filter out the missing values. Remember those are species that have not be identified as either not threatened, threatened, or extinct. In the right pane, I'll click the Filters icon. Then, I'll select New filter Threat Level and clear Include missing values. The automated explanation is updated to only view species that have been categorized as species of concern, not threatened, threatened, or extinct. Because the response variable for the automated explanation is a category data item, I can choose which event level to describe in the upper right corner of the chart. I'll choose Threatened. Now, I can maximize the automated explanation and view the explanation description tab. This tab describes the steps taken to determine the relevant factors, the most related factors, and the groups for threatened species. The Screening Results tab shows details about why certain data items were rejected as factors. The Relative Importance tab shows the relative importance score for each identified underlying factor where the most important factor is assigned a score of 1, and all other scores are proportional to that value. This information can also be viewed in the factors bar chart for the automated explanation object. When I'm finished viewing details, I'll restore my view. The automated explanation describes threat level in several ways. The characteristics of Threat Level show that a species has a 12.48% chance of being threatened, which is the second most common value for threat level. The factors bar chart shows the underlying factors for Threat Level. Notice, Order is the most important factor. I can click an underlying factor in the bar chart (for example, Category) and on the right side I can see the relationship between the Threat Level and the selected underlying factor. This relationship plot displays a stacked bar chart of the selected factor and each bar is split to show the portion of each level that represents threatened species and the portion that represents non-threatened species. Notice, Vascular plants, have a relatively large portion of threatened species. The groups list shows the top three and bottom three groups based on the selected factor (Category). Several decision trees are run on the response variable (Threat Level) and each group describes a leaf from one of the decision trees. The High tab lists the three groups that are likely to have the highest chance of being threatened and describes their characteristics, while the Low tab lists the three groups that are likely to have the lowest chance of being threatened and describes their characteristics. Let's say I want to use the characteristics from the highest group (82.08%) to try to identify species that are not currently threatened but belong to these groups. These would be species that fit the traits of other threatened species but aren't currently threatened. I may want to monitor these species to see if they become threatened in the future. To do this, I'll select the highest group, right-click the group and select Derive group item. A new calculated item is added to the Data pane and contains two distinct values: In if the species is in the group and Out if the species is not in the group. Then, I'll use this new data item with other data items to evaluate the species. I'll select Park Name and Common Names in the Data pane, right-click the data items and select Add to new page. A list table is added to a new page that shows the species present in each park. Remember, I want to determine the species that fall in the 82.08% chance group but are not currently classified as threatened. To make this list, I'll add a filter based on the derived group and clear Out (species not in the group) and Include missing values (species that were not identified). Then, I'll select the list table and on the Roles pane, I'll add Threat Level to the Columns role. Some species are already on the threatened list. I can click Threat Level in the list table to sort the list in ascending order and those species that have a high chance of being threatened, but are not yet classified as threatened, are displayed at the top of the list table.