20楼: Originally posted by tmdxyz at 2011-12-01 17:18:37:
y=p1+p2/Ln(x)+p3/(Ln(x))^2+p4/(Ln(x))^3+p5/(Ln(x))^4+p6/(Ln(x))^5+p7/(Ln(x))

Function: y = p1+p2/Ln(x)+p3/(Ln(x))^2+p4/(Ln(x))^3+p5/(Ln(x))^4+p6/(Ln(x))^5+p7/(Ln(x))^6+p8/(Ln(x))^7+p9/(Ln(x))^8+ ...

11楼: Originally posted by dbb627 at 2011-11-28 16:55:58:
[code]
x = [0.768;
0.711;
0.685;
0.671;
0.659;
0.657;
0.655;
0.654;
0.653;
0.654];
y = [4.584;
4.375;
4.234;
4.128;
4.041;
3.972;
3.913;
3.861;
3.815;
3.774];
opts = fitoptions ...

23楼: Originally posted by guojunhang at 2011-12-03 17:41:42:
程序运行完毕后
        a =       4.167
        b =      0.1139
我想检验这个拟合效果的时候，使用这样一段程序，
但是执行效果显示，很差，说明这个拟合结果严重不靠谱啊。
期待解释。
其实上面也有 ...

a =  4.167; b = 0.1139;
x = [0.768;
0.711;
0.685;
0.671;
0.659;
0.657;
0.655;
0.654;
0.653;
0.654];
y = [4.584;
4.375;
4.234;
4.128;
4.041;
3.972;
3.913;
3.861;
3.815;
3.774];
figure;
plot(x,y,'k-')
x1=x-mean(x);
x1=x1/std(x1);
hold on;
y_fit = a+b*log10(1-b*x1)+(1-b)*log10(1-b*x1+x1);
plot(x,y_fit,'ro-')

24楼: Originally posted by dbb627 at 2011-12-03 20:10:35:
这个x的数据做过归一化处理后再回归的
[code]
a =  4.167; b = 0.1139;
x = [0.768;
0.711;
0.685;
0.671;
0.659;
0.657;
0.655;
0.654;
0.653;
0.654];
y = [4.584;
4.375;
4.234;
4.128;
4. ...

25楼: Originally posted by guojunhang at 2011-12-03 21:39:31:
我抖胆再问一句：
a = ? ，  b =?

26楼: Originally posted by dbb627 at 2011-12-03 21:44:23:
a =  4.167; b = 0.1139;这个程序中不是给了吗，有什么不清楚的吗