请DINGDA老师帮忙!!
这里面有函数嵌套,不知我的输入是否表达有误,输入1STOPT里没有反映,请DINGDA老师帮忙!!谢谢啊!Title "00";
Parameters a,b,c,d;
Variable x,y1;
VarParameter y2,y3,y4,y5;
SharedModel;
Function
y1=(y3*x*0.56495325+y2*(0.549452555-x)*(0.02438585+0.025909*(1/(1+(x/a)^(-b*0.84276))))+y4*0.450547445*d)/(y3*x+y2*(0.549452555-x)+y4*0.450547445);
y2=(1/3)*(2/(1+((0.02438585+0.025909*(1/(1+(x/a)^(-b*0.84276))))/y5-1)*c)+1/(1+((0.02438585+0.025909*(1/(1+(x/a)^(-b*0.84276))))/y5-1)*(1-2*c)));
y3=(1/3)*(2/(1+(0.56495325/y5-1)*c)+1/(1+(0.56495325/y5-1)*(1-2*c)));
y4=(1/3)*(2/(1+(d/y5-1)*c)+1/(1+(d/y5-1)*(1-2*c)));
y5=0.02438585+0.025909*(1/(1+(x/a)^(-b*0.84276)))+(1/(1+(x/a)^(-b*0.84276)))*(0.56495325-(0.02438585+0.025909*(1/(1+(x/a)^(-b*0.84276)))));
data;
x,y1=
0.0620 0.288
0.1276 0.437
0.1742 0.532
0.2096 0.577
0.3286 0.74
0.3694 0.884
0.3927 0.972
0.4371 1.143
0.5253 1.223
0.5322 1.204
enddataset; 正确代码如下:
Title "00";
Parameters a,b,c,d;
ConstStr y5=0.02438585+0.025909*(1/(1+(x/a)^(-b*0.84276)))+(1/(1+(x/a)^(-b*0.84276)))*(0.56495325-(0.02438585+0.025909*(1/(1+(x/a)^(-b*0.84276))))),
y2=(1/3)*(2/(1+((0.02438585+0.025909*(1/(1+(x/a)^(-b*0.84276))))/y5-1)*c)+1/(1+((0.02438585+0.025909*(1/(1+(x/a)^(-b*0.84276))))/y5-1)*(1-2*c))),
y3=(1/3)*(2/(1+(0.56495325/y5-1)*c)+1/(1+(0.56495325/y5-1)*(1-2*c))),
y4=(1/3)*(2/(1+(d/y5-1)*c)+1/(1+(d/y5-1)*(1-2*c)));
Variable x,y1;
Function y1=(y3*x*0.56495325+y2*(0.549452555-x)*(0.02438585+0.025909*(1/(1+(x/a)^(-b*0.84276))))+y4*0.450547445*d)/(y3*x+y2*(0.549452555-x)+y4*0.450547445);
data;
//x,y1=
0.0620 0.288
0.1276 0.437
0.1742 0.532
0.2096 0.577
0.3286 0.74
0.3694 0.884
0.3927 0.972
0.4371 1.143
0.5253 1.223
0.5322 1.204
结果:
(RMSE): 0.0145181338831429
(SSE): 0.00210776211448863
(R): 0.999013104343221
(R^2): 0.99802718264948
---------- -------------
a 0.00931437175129268
b -2.1702014929257
c -0.0483359679641019
d 0.509969766796357 dingd 老师 ,为什么我的计算结果和你不一样呢?我用的都是默认的算法和设置每次计算的结果也不相同,结果差别挺大,我用的1.5注册版,是版本的原因吗?还是应该给参数设定合适的范围才能正确计算?
"00"
迭代数: 21
计算用时(时:分:秒:微秒): 00:00:51:766
优化算法: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
计算结束原因: 达到收敛判断标准
均方差(RMSE): 0.0279012042685422
残差平方和(SSE): 0.00778477199634917
相关系数(R): 0.996179760369045
相关系数之平方(R^2): 0.992374114968929
决定系数(DC): 0.992334050225161
卡方系数(Chi-Square): 0.00879228169472202
F统计(F-Statistic): 1041.05856400985
参数 最佳估算
---------- -------------
a 0.533218831963338
b -2.91167052961833
c 0.89603882971316
d -0.264388098152993
====== 结果输出 =====
No 实测值y1 计算值y1
1 0.288 0.3443463
2 0.437 0.4264247
3 0.532 0.4917516
4 0.577 0.5457608
5 0.74 0.7796007
6 0.884 0.8938094
7 0.972 0.9697738
8 1.143 1.1286227
9 1.223 1.2271294
10 1.204 1.2009908 这道题要到达2楼的结果还是很困难的,多算几次,我用4.0版,有大约一半的成功概率。(用其他软件,几乎无可能)。
页:
[1]