libname snail'D:\Temp';
data snail.rsrl;
Input T P Tc Pc Y / lackfit;
Cards;
358.885 5.242 1
1 161.0
358.885 1.758 1
-1 129.0
141.115 5.242 -1 1
166.0
141.115 1.758 -1 -1
135.0
375.0 3.5 -1.148 0
187.0
125 3.5 -1.148 0
170.0
250.0 5.5 0 1.148 174.0
250.0 1.5 0 -1.148 146
250.0 3.5 0 0 203.0
250.0 3.5 0 0 195.0
250.0 3.5 0 0 211.0
;
Proc RSreg;
Model Y=Tc Pc / lackfit;
Run;
Data grid;
Do;
Y=.;
Do T=110 to 390 by 20;
Do P=1.0 to 6.0 by .5;
Output;
End;
End;
End;
Data grid;
Set rsrl grid;
Proc RSreg data=grid out= predict noprint;
Model Y=T P / predict;
Run;
Data plot;
Set predict(firstobs=12);
Proc G3d data=plot;
Plot T*P=Y / yticknum=8 xticknum=6 zticknum=6 zmin=0 zmax=250
rotate=30 tilt=75 ctop=red cbottom=blue caxis=black;
Run;
Proc Gcontour data=plot;
Plot T*P=Y/ yticknum=6 xticknum=6 nlevels=10 caxis=black
vaxis=axis haxis=axis legend=legend autolabel;
axis value=(height=16pt) label=(height=16pt);
legend value=(height=16pt) label=(height=16pt);
Run;
Output:
The SAS System
16:46 Sunday, November 7, 2016
1
The RSREG
Procedure
Coding Coefficients for the
Independent Variables
Factor
Subtracted off
Divided by
Tc
-0.074000
1.074000
Pc
0.574000
0.574000
Response
Surface for Variable Y
Response Mean
178.200000
Root MSE
.
R-Square
1.0000
Coefficient of Variation
.
Type I Sum
Regression
DF
of Squares
R-Square F
Value Pr > F
Linear
2
881.008593
0.7629
.
.
Quadratic
1
273.788985
0.2371
.
.
Crossproduct
1
0.002422
0.0000
.
.
Total Model
4
1154.800000
1.0000
.
.
Sum of
Residual
DF
Squares
Mean Square F Value
Pr > F
Lack of Fit
0
0
.
.
.
Pure Error
0
0
.
Total Error
0
0
.
Parameter
Estimate
Standard
from Coded
Parameter
DF
Estimate
Error t
Value Pr > |t|
Data
Intercept
1
203.000000
.
.
.
188.604151
Tc
1
-2.408718
.
.
.
-0.379970
Pc
1
-25.261324
.
.
.
-14.496123
Tc*Tc
1
-14.238676
.
.
.
-16.423971
Pc*Tc
1
-0.091282
.
.
.
-0.056273
Pc*Pc
0
0
.
.
.
0
Sum
of
Factor
DF
Squares
Mean Square
F Value
Pr > F
Tc
3
278.164347
92.721449
.
.
Pc
2
737.297865
368.648933
.
.
The SAS System
16:46 Sunday, November 7, 2016
2
The RSREG
Procedure
Canonical Analysis of
Response Surface Based on Coded Data
Critical Value
Factor
Coded
Uncoded
Tc
-257.602125
-276.738682
Pc
150361
86308
Predicted
value at stationary point: -1089589
Eigenvectors
Eigenvalues
Tc
Pc
0.000048202
-0.001713
0.999999
-16.424019
0.999999
0.001713
Stationary point is a saddle
point.
The SAS System
16:46 Sunday, November 7, 2016
3
The RSREG
Procedure
Coding Coefficients for the
Independent Variables
Factor
Subtracted off
Divided by
Tc
-0.074000
1.074000
Pc
0.574000
0.574000
Response
Surface for Variable Y
Response Mean
178.200000
Root MSE
.
R-Square
1.0000
Coefficient of Variation
.
Type I Sum
Regression
DF
of Squares
R-Square F
Value Pr > F
Linear
2
881.008593
0.7629
.
.
Quadratic
1
273.788985
0.2371
.
.
Crossproduct
1
0.002422
0.0000
.
.
Total Model
4
1154.800000
1.0000
.
.
Sum of
Residual
DF
Squares
Mean Square F Value
Pr > F
Lack of Fit
0
0
.
.
.
Pure Error
0
0
.
Total Error
0
0
.
Parameter
Estimate
Standard
from Coded
Parameter
DF
Estimate
Error t
Value Pr > |t|
Data
Intercept
1
203.000000
.
.
.
188.604151
Tc
1
-2.408718
.
.
.
-0.379970
Pc
1
-25.261324
.
.
.
-14.496123
Tc*Tc
1
-14.238676
.
.
.
-16.423971
Pc*Tc
1
-0.091282
.
.
.
-0.056273
Pc*Pc
0
0
.
.
.
0
Sum
of
Factor
DF
Squares
Mean Square
F Value
Pr > F
Tc
3
278.164347
92.721449
.
.
Pc
2
737.297865
368.648933
.
.
The SAS System
16:46 Sunday, November 7, 2016
4
The RSREG
Procedure
Canonical Analysis of
Response Surface Based on Coded Data
Critical Value
Factor
Coded
Uncoded
Tc
-257.602125
-276.738682
Pc
150361
86308
Predicted
value at stationary point: -1089589
Eigenvectors
Eigenvalues
Tc
Pc
0.000048202
-0.001713
0.999999
-16.424019
0.999999
0.001713
Stationary point is a saddle
point.
加载中，请稍候......
加载中…
喜欢