SPSS data analysis-segmented regression

In the process of SPSS nonlinear regression, we talked about the loss function button can customize the loss function, but there is a constraint button is not mentioned, the function of the button is to self-
The parameter setting condition of the loss function is defined, these conditions are usually composed of the logical expression, which makes the loss function have certain judgment ability.

The main function of this function is to carry out piecewise regression, sometimes, the relationship between variables is not static, but has some kind of periodic characteristics, if we are independent of each stage
The fitting of the words, will cause more parameters, can not grasp the overall problem essence, and the sample is divided into multiple, will affect the accuracy, and the use of nonlinear regression definition
Constraints, you can use a model expression to fit each stage, thereby consolidating the stages to improve the accuracy of the model prediction

Look at a data, for the study of the relationship between temperature and power consumption, a set of data collected, first of all to do a scatter plot to see the relationship between the two sets of data

SPSS data analysis-segmented regression