Sample Data and Independent Variables
All regression methods require sample data derived from field measurements to train the regression model. The sample data can be imported using the “Import Training Data” dialog of each regression model. The imported file should be in text format (.txt extension) and must contain a header to label each column. The first two columns should be the X and Y coordinates of each sample, followed by any dependent values. Note that multiple dependent values can be given, but only one can be used for each regression execution. See Sample Data File in the File Formats section of the Appendix for information about sample data file.
- Sample Data: Click next to the sample data input box to select sample data. Imported values can then be viewed in the Import Training Data section.
- Dependent Variable: This parameter defines the dependent variable used in the regression analysis. Only one dependent variable from the sample data file can be selected each time.
- Plot Type: This parameter defines the plot type. It should be selected based on the plot surveying methodology.
- Square (default): The plot is square.
- Circle: The plot is circle.
- Length (meter)：When the plot type is square, set the length of the plot.
- Radius (meter)：When the plot type is circle, set the radius of the plot.
- Local Error Buffer：When the plot type is circle, set the radius of the plot.
- X: X coordinate of the plot center.
- Y: Y coordinate of the plot center.
For the Independent Variables section, users can select the independent variables from Elevation Percentile, Elevation Density, Intensity Percentile, LAI, Gap Fraction and Canopy Cover. Users can also import data in .csv or .tif format. While importing a .csv table, only one can be selected at a time, and must contain at least four fields, including X, Y, X size, and Y size (each column would be added to the list as an independent variable). However, multiple .tif images can be added at once. Every time a .tif image is added successfully, an independent variable named after the file name of .tif will be added to the list.
Note: The selected sample data should be within the extent of LiDAR point cloud. The XY range and scale of all imported .tif independent variables must be consistent. While importing a .csv file, X, Y, X size, and Y size will be imported as independent variables to the list. In this case, it is recommended to remove these 4 variables, or at least remove X size and Y size, depending on the case. It is also recommended to have more samples than dependent variables, otherwise it could lead to infinite solutions when solving the matrix.