Available with Image Analyst license.
The Predict Using Regression function computes a predicted raster based on raster data inputs and a regression model. The regression model is the output from the Train Random Trees Regression Model tool.
The regression model is defined in an Esri regression definition (.ecd) file. It contains all the information for a specific dataset or a set of datasets, and the regression model is generated by the Train Random Trees Regression Model tool.
The input can be a single band, a multiband, or a multidimensional raster, or a list of these types. The types of the input rasters must be the same type of raster trained by the regression model.
- When input is a multiband raster, each band is treated as a predictor variable. The bands must be in the same order as the multiband input for the regression model training tool.
- When input is a multidimensional raster, each variable is treated as a predictor variable, and the variable must be single band and have a time dimension. The variable order and names must be the same as the input when the regression model was trained. The output is a multidimensional raster.
- The input can be a list of items. The number of the items and the order of the items must match the input when the regression model was trained.
The following table describes the parameters:
The raster dataset or datasets representing the predictor variables. It can be a single-band raster, multiple-band raster, multidimensional raster, mosaic dataset, or a raster collection.
Input Definition File
The input Esri regression definition (.ecd) file that contains the statistics and information for the specific dataset, regression model, and chosen attributes.