作者waleic (深呼吸)
看板Statistics
标题[程式] SigmaPlot 回归分析之信赖区间
时间Sat May 10 23:00:19 2014
[软体程式类别]:
SigmaPlot
[程式问题]:
回归
[软体熟悉度]:
新手(不到1个月)
[问题叙述]:
目前使用了回归分析中的Logarithm,也顺利做出图;
欲请教的是,如果想知道95%信赖区间那两条曲线之方程式,
能使用什麽方法让软体计算出。
因为Report中只有显现所求回归分析之方程式,包含Y截距(y0)与常数项(a),
有关图上跑出之95% Confidence Band 及 95% Prediction Band,
没有看到任何相关系数。
虽然在Report中有跑出原本各个样本数值95% Confidence & Prediction 的 Row data,
是否仅能利用这些数据自行新列表计算?
[程式范例]:
Nonlinear Regression
Equation: Logarithm, 2 Parameter I
f=if(x>0, y0+a*ln(abs(x)), 0)
R Rsqr Adj Rsqr Standard Error of Estimate
0.7308 0.5341 0.5095 5.0703
Coefficient Std. Error t P VIF
y0 0.6651 4.5552 0.1460 0.8855 16.9499<
a 3.8810 0.8317 4.6666 0.0002 16.9499<
95% Confidence:
Row Predicted 95% Conf-L 95% Conf-U 95% Pred-L 95% Pred-U
1 25.4659 22.4863 28.4455 14.4434 36.4885
2 24.1809 21.5259 26.8358 13.2416 35.1201
3 25.5560 22.5509 28.5612 14.5265 36.5856
4 26.6084 23.2825 29.9344 15.4872 37.7296
5 9.6015 3.8722 15.3309 -2.4585 21.6616
6 25.5369 22.5372 28.5366 14.5089 36.5649
7 22.4357 20.0632 24.8083 11.5616 33.3099
8 19.3728 16.9033 21.8424 8.4771 30.2686
9 24.1444 21.4974 26.7914 13.2071 35.0818
10 22.8404 20.4219 25.2589 11.9561 33.7247
11 9.6015 3.8722 15.3309 -2.4585 21.6616
12 13.1577 8.8389 17.4766 1.7003 24.6151
13 23.9183 21.3189 26.5177 12.9923 34.8442
14 23.5947 21.0579 26.1314 12.6835 34.5059
15 23.7800 21.2082 26.3518 12.8606 34.6994
16 23.2875 20.8038 25.7713 12.3885 34.1865
17 22.9039 20.4771 25.3307 12.0177 33.7901
18 12.2917 7.6403 16.9431 0.7049 23.8785
19 22.4357 20.0632 24.8083 11.5616 33.3099
20 23.6157 21.0751 26.1563 12.7036 34.5278
21 22.6702 20.2726 25.0678 11.7905 33.5499
Fit Equation Description:
[Variables]
x = col(1)
y = col(2)
reciprocal_y = 1/abs(y)
reciprocal_ysquare = 1/y^2
'Automatic Initial Parameter Estimate Functions
F(q)=ape(ln(abs(x)),y,1,0,1)
[Parameters]
y0 = F(0)[1] ''Auto {{previous: 0.665099}}
a = F(0)[2] ''Auto {{previous: 3.88105}}
[Equation]
f=if(x>0, y0+a*ln(abs(x)), 0)
fit f to y
''fit f to y with weight reciprocal_y
''fit f to y with weight reciprocal_ysquare
[Constraints]
[Options]
tolerance=1e-10
stepsize=1
iterations=200
Number of Iterations Performed = 1
-------------
以上,感激不尽!
--
※ 发信站: 批踢踢实业坊(ptt.cc), 来自: 59.115.169.191
※ 文章网址: http://webptt.com/cn.aspx?n=bbs/Statistics/M.1399734025.A.D01.html