5. Spatial and temporal data subsetting

PAVE allows you to easily subset your data by geographic region, layer range, and time range. This section explains the concepts of how PAVE manages this information in its memory space. An understanding of this section should help you with your PAVE usage. To adjust layer ranges, geographic regions of interest (domains), or time step ranges for datasets and formulas, you can use menu items in the Datasets and Formulas menus on PAVE's main interface window.

Each dataset has some number of layers greater than or equal to one, some number of time steps greater than or equal to one, and some geographic region onto which its grid maps. A geographic region, often referred to as a domain, is defined by the map projection type (Lat/Lon, UTM, Lambert Conformal, etc.), the number of rows and columns in the dataset, and the geographic boundaries of the area on which the grid falls.

For each unique geographic region (or domain) that PAVE learns about by examining the dataset(s) chosen by the user, PAVE keeps a single domain object. Domain objects are where the currently selected geographic region for formulas and datasets are stored in memory. When a domain object is created by PAVE, it defaults to having all of its cells selected, until the user chooses a subregion within that domain. Because PAVE only stores a single domain object for each unique domain that it learns of, there may be numerous datasets and formulas that make use of the same domain object. Therefore, when a user chooses to modify a formula's or a dataset's selected region of interest, the region of interest for all formulas and datasets that refer to the same region and have the same grid dimensions will be affected. Any subsequent plots of any formulas who rely on that domain object will reflect its newly selected geographic region of interest.

Similarly, PAVE uses layers objects to store information about currently selected layer ranges. Each dataset i has some Ni number of layers associated with it, and each formula j has some Nj number of layers associated with it. For each unique number of layers that PAVE learns about, PAVE keeps a single layers object in memory. Therefore, when a user chooses to modify a formula's or a dataset's selected layer range, there may be more than one dataset and more than one formula on which the modified layer range has an effect. When PAVE learns about a new unique number of layers, the layers object that it creates by default has only the first layer selected, until the user modifies the layer range for this layers object.

Time ranges are handled differently than layer ranges and geographic regions of interest. Each dataset has exactly one unique time range object associated with it. That time range object is associated only with that one dataset. Each formula also has exactly one unique time range object associated with it. That time range object is associated only with that one formula. Because a formula may rely on more than one dataset to derive its data, a formula's time step range is always bounded by the time ranges of the datasets it depends on. For example, a formula could have three datasets from which it derives data. Those datasets could have 72, 48, and 24 time steps of data in their currently selected time range objects. In this case, the maximum number of time steps in that formula's time range object could ever have would be 24. Should the user reduce the number of steps in the currently selected time range of any of the three datasets to less than 24, then the maximum number of time steps in that formula's time range object would immediately be reduced accordingly.

Navigating through PAVE's menu items

Return To Table of Contents