Exercise 1. Sample Data
Regardless of the regression algorithm used, a sample dataset is always required to train the model. Training samples can be obtained through ground survey or base station/backpack laser scanning. A training sample should be a comma separated text file (.txt) that contains a descriptive header, and on each row, the XY coordinates of a sample plot followed by any number of dependent variables. Note that Regression Analysis allows multiple dependent variables in the training sample, but will consider only one of them in each regression model. In the example shown below (a section of the "SampleData.txt" file), the first two columns are the XY coordinates of circular sample plots of 10m radiuses, and the third contains a dependent variable, Biomass.