The task manager provides a framework for transforming and analyzing your data. By default, the Task Manager view first appears when you select a method from the "Method" pull down menu, and will appear with the first section (either "Input" or "Input parameters", as shown below, or "Regression Models" for regression methods) open. In the image below, the Task Manager view shows the input screen for the Disparity method. Note that the method name appears at the top, with an "x" that you can select to remove that method from the Task Manager view. Then, for most methods, the Task Manager view will contain four sections, labeled "Input", "Advanced" (sometimes not present), "Output", and "Run Method" (see green circles in the image below). The sections for regression methods are different, and are shown here.
As the name suggests, the "Input" page is where you identify the datasets that you would like to work with, relevant times, and other factors that will determine how STIS performs the data transformation or statistical analysis. The "Advanced" page typically provides an opportunity to adjust additional features, such as the p-value, or whether to use a Simes Correction. The "Output" page is where you can specify the name of output files, and the "Run method" page summarizes all of the other pages, and contains the "Run" button to implement the method.
You can link directly to guidance on what to enter in each of these sections by clicking on the following methods: Z-Score; Aggregation; Empirical Bayesian Smoothing; Local G; Local G*; Univariate Local Moran; Bivariate Local Moran; Turnbull; Besag and Newell, Disparity, Aspatial Regression, and Geographically-weighted Regression.
Encompassing the methods within the task manager framework has several advantages. First, by putting all of the methods into a similar framework, we hope to simplify the process of learning how to use the many powerful tools in STIS. In addition, because the task manager works as a view, you can hide and show it without losing the information that you entered for a particular analysis -- this information is even retained if you close a project. You can run and re-run several versions of the method (i.e., with different datasets or different "Advanced" settings) by selecting the method again from the pull-down menu (see the image below, where the "Disparity" method appears twice), or can just edit the pages for a given method to re-run an analysis using different options (then only the last version will be saved). Finally, you can run different methods, and the task manager will maintain this information until you choose to "close" a method by clicking on the "x" to the left of the method name. When you have run several methods and the "Method Name" bar (shown in circled in green, below) fills up, a pair of arrows will appear in the top right part of the bar that you can use to scroll through the methods you have used.
Depending on the screen space available when you invoke a method, the task manager may open in a "compressed" version that does not show you the entire screen for a a particular section (e.g., "Input") at the same time. Note that if you can see horizontal or vertical "sliders" in a given section of the Task Manager (shown circled in red for the Input page of the Task Manager View of the Univariate Local Moran, below left), then you will need to expand the size of the Task Manager View by clicking on the window edges and stretching it, or you will need to use the scrollbars to see all of the relevant information.
|
Compressed version |
All of the "Input" window shown |
|
|
|
You can "undock" the Task Manager
window by clicking on the 
symbol in the upper right corner, and hide
it by clicking on the 
button. Undocking allows you to move the
window around; if you choose to close the window, you can view it again
by checking the box next to "Task Manager view" in the "Window"
menu.
