Simultaneous Equations - Three Variables
Codewars Kata 52√
Description
https://www.codewars.com/kata/59280c056d6c5a74ca000149/solutions/python
-简述：本题求解三元一次方程，输入为3个列表为各个方程的参数，输出一个列表为三个根的解。
-思路：用np.linalg.solve求解线性方程组，根据要求变换列表的形式。
-难点：如何用最少的语句变换参数形式
We have 3 equations with 3 unknowns x, y, and z and we are to solve for these unknowns.
Equations 4x -3y +z = -10, 2x +y +3z = 0, and -x +2y -5z = 17 will be passed in as an array of [[4, -3, 1, -10], [2, 1, 3, 0], [-1, 2, -5, 17]] and the result should be returned as an array like [1, 4, -2] (i.e. [x, y, z]).
My solution
import numpy as np
def solve_eq(eq):
lst1 = []
lst2 = []
for i in range(0,len(eq)):
lst1.append(eq[i][-1])
lst2.append(eq[i][:-1])
rst = np.linalg.solve(lst2, lst1)
return [int(round(x)) for x in list(rst)]
Given solution
import numpy as np
def solve_eq(eq):
a = np.array([arr[:3] for arr in eq])
b = np.array([arr[-1] for arr in eq])
return [round(x) for x in np.linalg.solve(a,b)]
Improved solution
import numpy as np
def solve_eq(eq):
lst1 = [arr[:-1] for arr in eq]
lst2 = [arr[-1] for arr in eq]
rst = np.linalg.solve(lst1,lst2)
return [int(round(x)) for x in list(rst)]
Points:
1 解形如AX=b的线性方程组：np.linalg.solve(A,b)
2 参考答案中用的np.array会报错，去掉之后直接用[]反而可以运行。
