本文介绍了一个能用最小二乘法原理自动地回归一个一元线性方程或一元非线性方程的程序 如果将一组实验数据[(x_i,v_i),i=1,2,…,n]送入计算机内,该程序将自动地从下列给定的方程中选出一个最佳的回归方程: 1、y=a+bx, 2、y=ax~b, 3、y=ae~(bx), 4、y=a+blgx, 5、y=ae~(b/x), 6、1/y=a+b/x, 7、y=1/(ax+b), 8、y=1/(a+be~(-x)), 9、y=ax~+C, 10、y=ae~(bx)+C。 In this paper, we introduce a procedure that can automatically return a univariate linear equation or a univariate nonlinear equation using the principle of least squares. If a set of experimental data [(x_i, v_i), i = 1,2, ..., n] , The program automatically selects the best regression equation from the given equations given below: 1, y = a + bx, 2, y = ax ~ b, 3, y = ae ~ , Y = a + blgx, 5, y = ae ~ b / x, 6,1 / y = a + b / x, 7, y = 1 / ax + b, y = 1 / a + be ~ (-x)), 9, y = ax ~ + C, 10, y = ae ~ (bx) + C.