numpy. linalg. Lstsq (a, B, rcond=-1) parsing

numpy.linalg.lstsq(a,b,rcond=-1) analysis

Returns the least square solution of a linear matrix , The o x Go to the fitting $ax=b$. Minimize Euclid by calculation 2 norm $||b-ax|{|}^2$ Vector $x$

• Parameters ：a:{M,N} coefficient matrix
• Parameters ：b{(M,),(M,K)} Ordinate or dependent variable value . If b It's two-dimensional , Then for b Of K Calculate the least squares solution for each of the columns
• Parameters ：rcond Floating point numbers , Optional a Cut off ratio of small outliers of （ It is about controlling the number of outliers ）.
• Return value ：X{(N,),(N,K)} Solution of least square method . When b The dimension of is (M,) when ,X The dimension of is (N,), When b The dimension of is (M,K) when ,X The dimension of is （N,K）.