各位大侠请帮助小弟:怎样用一个周期的正弦函数去拟合40多

见下例：
x=[-.3000, 0., .3000, .6000, .9000, 1.200, 1.500, 1.800, 2.100, 2.400, 2.700, 3., 3.300, 3.600, 3.900, 4.200, 4.500, 4.800, 5.100, 5.400, 5.700, 6., 6.300, 6.600, 6.900, 7.200, 7.500, 7.800, 8.100, 8.400, 8.700, 9., 9.300, 9.600, 9.900, 10.20, 10.50, 10.80, 11.10, 11.40, 11.70, 12., 12.30]
y=[-.1853, .2964, .3068, .8302, 1.057, 1.171, 1.027, 1.052, .9638, .8794, .4683, .3575, -.1257, -.2464, -.5395, -.6379, -.7630, -.7250, -.6585, -.6725, -.3411, -.2201, .2598e-1, .5348, .7284, .9376, 1.209, 1.182, 1.155, 1.112, .9046, .5851, .1793, -.1023, -.1916, -.6913, -.7327, -.9306, -.7009, -.7055, -.6118, -.3952, -.2453]
>> fx=@(b,x)b(1)+b(2)*sin(b(3)*x+b(4))
fx =
@(b,x)b(1)+b(2)*sin(b(3)*x+b(4))
>> b=rand(1,4)
b =
0.4168 0.6569 0.6280 0.2920
>> for l=1:3
b=lsqcurvefit(fx,b,x,y)
b=nlinfit(x,y,fx,b)
end
b =
0.1581 -0.9982 0.9957 -3.1115
b =
0.1581 -0.9982 0.9957 -3.1114
b =
0.1581 -0.9982 0.9957 -3.1114
b =
0.1581 -0.9982 0.9957 -3.1114
b =
0.1581 -0.9982 0.9957 -3.1114
b =
0.1581 -0.9982 0.9957 -3.1114
>> plot(x,y,'o')
>> hold on
>> y1=b(1)+b(2)*sin(b(3)*x+b(4));%or y1=fx(b,x);
>> plot(x,y1)
>> axis tight

你可以先设定一个正弦函数如：y=a*sin(b*x);
然后你可以用lsqcurvefit这个函数去拟合出这两个参数a和b，具体用法自己研究一下。

这位大哥能与小妹联系下吗，想请您在数据拟合方面帮个忙呢！

我想要拟合一曲线，原始数据是：
yt=[ 521.486755 161.973785 130.258820 120.327599 90.597206 51.033958 29.523090 19.248325 11.156927 10.221434 4.437991 3.455873 2.075573 1.551129 1.744870 0.761403 0.754921 0.532298 0.324880 0.587357 0.236039 0.227544 0.132627 2.709515]
x=0.2:0.2:4.8
想要把它拟合成自定义的函数形式：
（1）M-P分布： yt=a*exp(-b*x)
（2）Gamma分布： yt=a*x^b*exp(-c*x)
想写出m文件画出这两条拟合曲线，并且在图上给出拟合公式，不知道拟合后的相关系数能不能也做出来啊？在做的时候总是遇到很多的问题，所以想麻烦您帮忙下咯

clear,clc
x=0.2:0.2:4.8;
y=[ 521.486755 161.973785 130.258820 120.327599 90.597206 51.033958 29.523090 19.248325 11.156927 10.221434 4.437991 3.455873 2.075573 1.551129 1.744870 0.761403 0.754921 0.532298 0.324880 0.587357 0.236039 0.227544 0.132627 2.709515];
x=x(;y=y(;
SSy=var(y)*(length(y)-1)
fx1=@(b,x)b(1)*exp(-b(2)*x);
fx2=@(b,x)b(1)*x.^b(2).*exp(-b(3)*x);
b=rand(1,2)
for l=1:5
b=lsqcurvefit(fx1,b,x,y)
b=nlinfit(x,y,fx1,b)
end
figure(1)
y1=fx1(b,x);
RSS=(y-y1)'*(y-y1)
R12=(SSy-RSS)/SSy
hold off
plot(x,y,'o')
x1=0:.1:5;
y1=fx1(b,x1);
hold on
plot(x1,y1)
b=rand(1,3)
for l=1:5
b=lsqcurvefit(fx2,b,x,y)
b=nlinfit(x,y,fx2,b)
end
y2=fx2(b,x);
RSS=(y-y2)'*(y-y2)
R22=(SSy-RSS)/SSy
figure(2)
hold off
plot(x,y,'o')
y1=fx2(b,x1);
hold on
plot(x1,y1)

ding