而i些adMiodais►
10:
备品支出p
10
家庭收入1食品支出1
var
var
var
1
20.00
30.00
33.00
40.00
15.00
13.00
26.00
3B.0D
35.00
43.00
7.00
9.00
6.00
T1.00
5.00
4.00
6.00
10.00
9.00
2
3
4
RuEre:
5,ion
Lift*"--.[
5
LgglLineii-*
Rat*Raduetiaxt33-►
Jffonp-aretricTests.
Ti_meSeiries"
SuJTVl.v-ll.口
fflultipl«R♦后pon与q归
Nii£3-1nsJ£alu.erfkjialy£i-3...
CompieKSamples.
QurireEstima.t.1on...
6
Bin-ajryLo^ictic...
Mult*1JwiEtig..
Orjjiri-aJL--■Probit...
7
8
9
10
10.00
KonJ-inaur..
ghtl?
Etim4itian.:
£=SLaEsLuam七Sq.u.4res...
15
1
2(00
7.03
2
3C00
9.03
3
3:
00
8.03
4
4C.00
11.0J
5
1E,00
5.0J
J
1-00
4,0]
2
2E00
8.0J
1
3E00
10.0J
3E00
9.03
OK
He5el
hdepertferJfs!
母冢就入
SdectaiVdicUe:
□
Help
£«eLab5l5:
□
Reside
WLSWeghi:
□
$!
cndc(dcevatcns
22
■LinearRegression
LinearRegression:
Statistics
|Gjnl泣eCaicd
RegressionCce;lcenlS
PEstmatesv^CcttifercBHen,覆-CovaiiartemalriK
17Modelfit
rRscwiedchange
rQesairfves
rPalcndpailialccfrefetnj
FCojlrwiMd叨烟6“
Cepjrxfenl:
Blcckloll
24
16
V
11
5就比⑶JP&怆.|Save...
-Dyibin-Watson
r£&附加(1即而
什。
跳节c而佗(3
+Regression
VariablesEnteredjRernovedb
Model
Variables
Entered
Variables
Removed
Method
1
家庭I攵入i
•
Enter
a.Allrequestedvariablesentered.
b.DependentVariable:
便品支出
ModelSummary
Model
R
RSquare
AdjustedRSquare
Std.ErroroftheEstimate
1
.951a
904
892
.73304
aPredictors:
(Constant),家庭收入
ANOVAb
Model
Sumof
Squares
df
MeanSquare
F
Sig.
1Regression
40.601
1
40.601
75.559
.0003
Residual
4.299
8
.537
Total
44.900
9
a.Predictors:
Constant),家庭收入
b.DependentVariable:
食品支出
CoefficienlsB
Model
Mndardlzed
Coefficients
Standardized
Coefficients
t
Sig.
95%ConfidenceIntervalforB
B
Sid.Error
Beta
LowerBound
upperBound
1位喇
2.173
720
3.017
.512
3.333
家跳人
.202
023
-i:
i
8.692
000
149
.266
a.DependentVariable
判别:
?
=4十?
=2173+0.202(,且x与y的线性相关系数为R=0.951
回归方程的F检验值为75.559,对应F值的显著性概率是0.000<0.05,表示线性回归方程具有显著性,当对应F值的显著性概率>0.05,表示回归方程不具有显著性。
每个系数的t检验值分别是3.017与8.692,对应的检验显著性概率
分别为:
0.017(<0.05)和0.000(<0.05),即否定H0,也就是线性假设是显
著的。
二、一元非线性回归
例4,3.1出钢时所用的盛钢水的钢包,由于钢水对耐火材料的侵
蚀,容积不断增大.希望找出使用次数X与增大的容积丫之间的关系,试验数据列于表4.3」.
寰4.3J使用次数X与增大的容积Y的试验数据
若
2
3
4
5
6
7
8
9
6,42
8.20
9.58
9.50
9.70
10.00
9.93
9.99
苍
10
tl
12
13
14
15
16
y)
10.49
10.59
10.60
10.80
10.60
10.90
10.76
SPSSt解过程:
0Q|W|吧I国I“I年同函垂问3效
IS:
喜审*IO.TS
次数又
容枳VI
var|
var
1
2
6.42
2
3
3.20
3
4
9.58
4
5
9.60
5
6
9.70
6
7
10.00
7
8
9.93
e
9
9.99
9
1O
1049
10
11
1O59
11
12
1060
12
13
1O6Q
13
14
10.60
14
15
IO90
15
16
10.7^
16
1、Y与X的二次及三次多项式拟合:
谢。
昌
罔
-
3
c?
R卷ROUt£
tiveStatisticsk
*
Httnx►
LinearModel*
odels卜
It4
司
H
Descrip
Tabies
tr・
General
MixedM
15~
I^EK
1S
次数X
容积¥|
UI
var
1
2
6.42
r
1.8594
2
3
B.20
L
2.1041
3
d
9.55
4
5
950
|Regress
ion
置
Linear...
5
6
9.70
Loglinear>
Classifx►
DataReductioil上
Scale>
genpaarMttrieTtsts>
TimeSeries量
Swrvivil>
MjjltipleRe印onse卜
MissinzValueAnal^siz...
Cm-veEstimation...
—
6
7
10.00
BinaryLogiitia…
7
8
9.93
MultiziomialLogistic...
Ordinal.,.
Frobi匕…
B
9
9.99
g
-10
10.49
10
11
1059
Nonlinear...
iSightEstirnftti^n..r.
2-StsgeLeai^tSquired..-
11
12
10EO
12
13
10.80
13
14
10EO
Compleac
Samples・
OptimalScaling...
14
15
1090
IU.UUULJ
.3Ir
15
16
10.76
160000
□929
2.375B
17
18
19
Analyze
FileEditViewDataTraitsform
GraphsUtilitiesWindowHelp
次数*
容积丫
1
2
6.42
2
3
8.30
3
4
S.5S
4
5
9.60
5
6
970
6
7
1000
7
3
9.93
8
9
9.99
9
1U
10.49
10
11
10.50
11
12
10.60
12
13
1030
13
14
1060
14
15
10.90
15
ie
10.76
20
■CurtreEst]aation
VUV1B管您④合
DependentMthRsqd.f.
容弗只¥容枳丫
QUA.859
CUB.921
Q印enderM就
#Ymiatfe:
□元
「Tjffle
CaseLabels:
m।
Models
「Linear*Quadratic
一LogarithmicF叁Mg
厂l^veiseI-Power
F口中国《MQVA3*
EastsBesetCahcel
Help
BIncludeconstantinequator
RPlatrnocfefe
广Coflcoi^id厂Growth
广£厂E*ciEntial
lLogistic
U即rbmnd||
Save...
FSigfbOb1b2b3
1236.64□□□6092774OB-.0290
1142.B4□□□411B31.7DEB-.1534.口口4£
□OY
ObservedQuadkmtic:
Cul.dc
2
所以,二次式为:
Y=6.09270.7408x-0.029x
2_3
三次式为:
Y=4.1181.7068x一0.1534x20.0046x3
2、把Y与X的关系用双曲线拟合:
作双曲线变换:
U
Lv
次数XI
容积Y
VI
u
var
var
v
1
2
6.42
.5000
.1558
2
3
8.20
3330
.1220
3
4
9.58
.2500
.1040
4
5
9.50
.2000
.1053
5
6
9.70
.1667
.1031
6
7
10.00
.1420
.1000
7
8
9.93
.1250
.1007
8
9
9.99
.1110
.1001
9
10
10.49
.1000
0953
_10
11
10.59
.0900
.0944
11
12^
10.60
.0830
.0943
12
13
10.80
.0760
.0926
13
14
10.60
.0714
.0943
14
15
10.90
.0667
.0917
15
16
10.76
.0625
.0926
19
20
rQ
第叟]「卜
ICompute...
ll
n
►
VisualBander...
次数X
Count...
RankCases...
AutomaticRecode..
V1
U1
var
var
1
_2
2
3
1.8594
2.1041
_3
4
CreateTimeSeriesReplaceMissingRandomNumberSeed
...1
22SA7
_4
5
lues...
2.2513
_5
6
2.2721
6
7
2.3026
7
8
Kunre
2.2956
_8
9
9.99
9.0000
.1001
2.3016
9
10
10.49
10,0000
.0953
2.3504
10
11
10.59
11,0000
0944
2.3599
11
12
10.60
12.0000
0943
2.3609
12
13
10.80
13.0000
.0926
2.3795
13
14
10.60
14.0000
.0943
2.3609
14
15
10.90
15,0000
.0917
2.3888
15
16
10.76
16.0000
.0929
2.3758
Transform
eEditViewRata
AnalyzegraphsUtilitiesWincowHelp
17
18
19
次数X
J
容积Y]
■CoaputeVariable
1
2
6.42
TargetVariable:
NumericExpression
2
3
8.20
M
1/次数x
3
J
4
9.58
950
Type&LabeL|
6
7
970
1000
或容积Y
LU
+II7|8|9|Functions:
||
7
8
9
10
9.93
9.99
1049
V
U磅VI,》U1
」土I山13/业皿
ABS(numexpr)a
ANYJtest,value,value,..)
ARSIN(numexpr)
ARTAN(numexpr)
CDFNORM(zvalue)
COF.BERNOULU(q,p)▼
10
11
10.59
11
J2
12
13
1060
1080
H1
13
~I4
14
15
10.60
10.90
I
OK|Raste|Reset|Cancel|Help|
15
16
1D7K
・[・~・]
■一■一■一■■—一・
■■■■一」
1
4Regression
VariablesEnteredjRemovedb
Model
VariablesEntered
Variables
Removed
Method
1
Va
•
Enter
a.Allrequestedvariablesentered.
b.DependentVariable:
U
ModelSummary
Model
R
RSquare
AdjustedRSquare
Std.ErroroftheEstimate
1
.968』
.938
.933
.0042568
a.Predictors:
(Constant),V
ANOVAb
Model
SumofSquares
df
MeanSquare
F
Sig.
Regression
.004
1
.004
196.227
.000a
Residual
.000
13
.000
Total
.004
14
a.Predictors:
(Constant),V
b.DependentVariable:
U
coemuients0
Model
unslandardizedCoefficients
standardizedCoefficients
t
95%ConfidenceIntervalforB
B
Std.Error
Esta
Low&rBound
UpperBound
1(Conalarrt)
.082
.002
44,514
.000
.070
„086
V
131
009
963
14.DOB
.000
111
151
a.DependsmOanatue'U
11
判别:
U=0.082—0.131V,u=-,V=-,V与U的相关系数为R=0.968,回归yx
方程系数的F检验值为196.227,对应F值的显著性概率是0.000(<0.05),表示线性回归方程具有显著性,每个系数的t检验值分别是440514与14.008,
对应的检验显著性概率分别为:
0.000(<0.05)和0.000(<0.05),即否定H0,也就是线性假设是显著的。
Y…1
3、把Y与X的关系用倒指数函数拟合:
Y=aex,则lnY=lna+b
x
令U1=LN(Y),V1=V=1/x,有U1=c+bV1.
我数X
容积Y||
1
2
6.42
2
3
8.20]
3
4
9&
_4
5
9.5C
_5
6
9.70
5
7
10.0C
_7
9
9.93
_8
9
9.99
9
io
10.J9
W
iT
10.5S
1?
12
10.60
12
13
1Q.8C
13
14
10.60
U
15
10.90
15
16
10.76
MlCoiputeVariable
16
TargetVaiiable:
TypeJ;Label..
NumericExpression
ABSfnunnexprl酬1悔虬丫目电丫疝/一.)AFISIN上廊即。
AFlTAN[nwmej(prjCDFNORM(zvaluB)CDF.BERNOLJLU(q,p)
K/
4
>-(-
1:
".
次数X
容积Y
u
r
1
21
6.42
5000
.1558
2
3
8.20
.3333
.1220
3
4
9.58
2500
1044
4
5
9.50
2000
1053
5
6
9.70
1667
.1031
6
7
10.00
1429
1000
7
8
9.93
1250
1007
8
9
9.99
.1111
.1001
9
10
10.49
.1000
.0953
10
11
10.59
0909
0944
11
12
1060