用r软件做分位数回归.docx

用r软件做分位数回归.docx

《用r软件做分位数回归.docx》由会员分享，可在线阅读，更多相关《用r软件做分位数回归.docx（9页珍藏版）》请在冰豆网上搜索。

用r软件做分位数回归

分位数回归R〔quantreg〕

一、

假设你能能上国外的点击菜单上“程序包〞——>“安装程序包〞——>选择一个网站——>点击“quantreg〞就能下载并安装。

假设你不能上网站，把我附件中的文件解压后，放入R\library的子目录下就可以了。

二、

用library（quantreg）调入，就可以作分位数回归了。

下面的命令帮助你熟悉quantreg.

>help（package="quantreg"）

>help（rq）

三、下面是一个例子，或许对你有帮助。

data（engel）

engel

xy

1420.1577255.8394

2541.4117310.9587

3901.1575485.6800

4639.0802402.9974

5750.8756495.5608

6945.7989633.7978

7829.3979630.7566

8979.1648700.4409

91309.8789830.9586

101492.3987815.3602

11502.8390338.0014

12616.7168412.3613

13790.9225520.0006

14555.8786452.4015

15713.4412512.7201

16838.7561658.8395

17535.0766392.5995

18596.4408443.5586

19924.5619640.1164

20487.7583333.8394

21692.6397466.9583

22997.8770543.3969

23506.9995317.7198

24654.1587424.3209

25933.9193518.9617

26433.6813338.0014

27587.5962419.6412

28896.4746476.3200

29454.4782386.3602

30584.9989423.2783

31800.7990503.3572

32502.4369354.6389

33713.5197497.3182

34906.0006588.5195

35880.5969654.5971

36796.8289550.7274

37854.8791528.3770

381167.3716640.4813

39523.8000401.3204

40670.7792435.9990

41377.0584276.5606

42851.5430588.3488

431121.0937664.1978

44625.5179444.8602

45805.5377462.8995

46558.5812377.7792

47884.4005553.1504

481257.4989810.8962

492051.17891067.9541

501466.33301049.8788

51730.0989522.7012

52800.7990572.0807

531245.6964907.3969

541201.0002811.5776

55634.4002427.7975

56956.2315649.9985

571148.6010860.6002

581768.82361143.4211

592822.53302032.6792

60922.3548590.6183

612293.19201570.3911

62627.4726483.4800

63889.9809600.4804

641162.2000696.2021

651197.0794774.7962

66530.7972390.5984

671142.1526612.5619

681088.0039708.7622

69484.6612296.9192

701536.02011071.4627

71678.8974496.5976

72671.8802503.3974

73690.4683357.6411

74860.6948430.3376

75873.3095624.6990

76894.4598582.5413

771148.6470580.2215

78926.8762543.8807

79839.0414588.6372

80829.4974627.9999

811264.0043712.1012

821937.9771968.3949

83698.8317482.5816

84920.4199593.1694

851897.57111033.5658

86891.6824693.6795

87889.6784693.6795

881221.4818761.2791

89544.5991361.3981

901031.4491628.4522

911462.9497771.4486

92830.4353757.1187

93975.0415821.5970

941337.99831022.3202

95867.6427679.4407

96725.7459538.7491

97989.0056679.9981

981525.0005977.0033

99672.1960561.2021

100923.3977728.3997

101472.3215372.3186

102590.7601361.5210

103940.9218517.9196

104643.3571459.8177

1052551.6615863.9199

1061795.3226831.4407

1071165.7734534.7610

108815.6212392.0502

1091264.2066934.9752

1101095.4056813.3081

111447.4479263.7100

1121178.9742769.0838

113975.8023630.5863

1141017.8522645.9874

115423.8798319.5584

116558.7767348.4518

117943.2487614.5068

1181348.3002662.0096

1192340.61741504.3708

120587.1792406.2180

1211540.9741692.1689

1221115.8481588.1371

1231044.6843511.2609

1241389.7929700.5600

1252497.78601301.1451

1261585.3809879.0660

1271862.0438912.8851

1282021.85461509.7812

129697.3099484.0605

130571.2517399.6703

131598.3465444.1001

132461.0977248.8101

133977.1107527.8014

134883.9849500.6313

135718.3594436.8107

136543.8971374.7990

1371587.3480726.3921

1384957.81301827.2000

139969.6838523.4911

140419.9980334.9998

141561.9990473.2021

142689.5988581.2029

1431398.5203929.7540

144820.8168591.1974

145875.1716637.5483

1461392.4499674.9509

1471256.3174776.7589

1481362.8590959.5170

1491999.25521250.9643

1501209.4730737.8201

1511125.0356810.6772

1521827.4010983.0009

1531014.1540708.8968

154880.3944633.1200

1552432.39101424.8047

1561177.8547830.9586

1571222.5939925.5795

1581519.58111162.0024

159687.6638383.4580

160953.1192621.1173

161953.1192621.1173

162953.1192621.1173

163939.0418548.6002

1641283.4025745.2353

1651511.5789837.8005

1661342.5821795.3402

167511.7980418.5976

168689.7988508.7974

1691532.3074883.2780

1701056.0808742.5276

171387.3195242.3202

172387.3195242.3202

173410.9987266.0010

174832.7554614.7588

175614.9986385.3184

176887.4658515.6200

1771024.8177708.4787

1781006.4353734.2356

179726.0000433.0010

180494.4174327.4188

181748.6413429.0399

182987.6417619.6408

183788.0961400.7990

184831.7983620.8006

1851139.4945819.9964

186507.5169360.8780

187576.1972395.7608

188696.5991442.0001

189650.8180404.0384

190949.5802670.7993

191497.1193297.5702

192570.1674353.4882

193724.7306383.9376

194408.3399284.8008

195638.6713431.1000

1961225.7890801.3518

197715.3701448.4513

198800.4708577.9111

199975.5974570.5210

2001613.7565865.3205

201608.5019444.5578

202958.6634680.4198

203835.9426576.2779

204873.7375631.7982

205951.4432608.6419

206473.0022300.9999

207601.0030377.9984

208713.9979397.0015

209829.2984588.5195

210959.7953681.7616

2111212.9613807.3603

212958.8743696.8011

2131129.4431811.1962

2141943.04191305.7201

215539.6388442.0001

216463.5990353.6013

217562.6400468.0008

218736.7584526.7573

2191415.4461890.2390

2202208.78971318.8033

221636.0009331.0005

222759.4010416.4015

2231078.8382596.8406

224499.7510408.4992

2251020.0225775.0209

2261595.16111138.1620

227776.5958485.5198

2281230.9235772.7611

2291807.9520993.9630

230415.4407305.4390

231440.5174306.5191

232541.2006299.1993

233581.3599468.0008

234743.0772522.6019

2351057.6767750.3202

假设我们希望得到med[Y|X=x]=a+bx中位数回归，那么你可以用下面的命令：

fit1<-rq（y~x,tau=0.5,data=engel）

fit1中存放了你想得到的参数估计，因此你可直接用fit1调用：

fit1

Call:

rq（formula=y~x,tau=0.5,data=engel）

命令summary（fit1）会给你更多的结果：

Call:

rq（formula=y~x,tau=0.5,data=engel）

tau:

[1]0.5

Coefficients:

coefficientslowerbdupperbd

（Intercept）81.4822553.25915114.01156

x0.560180.487020.60199

用下面的命令可以得到残差的值：

r1<-resid（fit1）

r1

这是最简单，假设你想知道更详细的可查阅R网站上关于quantileregression的内容。