最优化方法与理论这本书的代码中有很多印刷错误的地方，大家平时要留意下复制最优化理论书上的源代码，用Matlab编译时出现如下错误：??? Error using == qmin Too many output arguments.工具/原料moreMatlab方法/步骤1/6分步阅读

书上的代码如下：

function = qmin (phi, a, b, delta, epsilon)

s0 = 1; maxj = 20; maxk = 30; big = 1e6; err = 1; k = 1;

S(k) = s0; cond = 0; h = 1; ds = 0.00001;

if (abs (s0) 1e4), h = abs(s0) * (1e-4); end

while (k maxk err epsilon cond ~= 5)

f1 = (feval (phi, s0 + ds) - feval(phi, s0 - ds)) / (2 * ds)

if(f1 0), h = -abs(h); end

s1 = s0 + h; s2 = s0 + 2 * h; bars = s0;

phi0 = feval(phi, s0); phi1 = feval (phi, s1);

phi2 = feval(phi, s2); barphi = phi0; cond = 0;

j = 0;

while (j maxj abs (h) delta cond == 0)

if (phi0 = phi1),

s2 = s1; phi2 = phi1; h = 0.5 * h;

s1 = s0 + h; phi1 = feval (phi, s1);

else if (phi2 phi1),

s1 = s2; phi1 = phi2; h = 2 * h;

s2 = s0 + 2 * h; phi2 = feval (phi, 2);

else

cond = -1;

end

end

j = j + 1;

if (abs (h) big || abs (s0) big), cond = 5; end

end

if (cond == 5),

bars = s1; barphi = feval (phi, s1);

else

d = 2 * (2 * phi1 - phi0 - phi2);

if (d 0),

barh = h * (4 * phi1 - 3 * phi0 - phi2)/d;

else

barh = h / 3; cond = 4;

end

bars = s0 + barh; barphi = feval (phi, bars);

h = abs (h); h0 = abs (barh);

h1 = abs (barh - h); h2 = abs (barh - 2 * h);

if (h0 h), h = h0; end

if (h1 h), h = h1; end

if (h2 h), h = h2; end

if(h == 0), h = barh; end

if (h delta), cond = 1; end

if (abs (h) big || abs (bars) big), cond = 5; end

err = abs (phi1 - barphi);

s0 = bars; k = k+1; S(k) = s0;

end

if (cond == 2 h delta), cond = 3; end

end

s = s0; phis = feval (phi, s);

ds = h; dphi = err;

end

2/6

在Matlab中新建一个M文件，名称为qmin

3/6

编译时出现如下错误提示：

大意是输出参数太多

4/6

仔细检查一下代码，发现命令语句的输出参数中多了一个k参数

5/6

将命令语句中的K参数删除，再次运行便可成功输出结果了

6/6

注意：最优化方法与理论这本书的代码中有很多印刷错误的地方，大家平时要留意下

注意事项

如果对本经验有什么不清楚的地方或遇到什么问题，可能是我描述的问题，您可以给我评论指教哦。我会及时给予回复的。

matlabmany