虚拟变量线回归源程序代码.doc

源代码显示原始数据：Private Sub Command1_Click()Open C:x.txt For Input As #1Open c:y.txt For Input As #2Dim X(1 To 22), Y(1 To 22), X1!, Y1!, a1%Do While Not EOF(1) Input #1, X1 a1 = a1 + 1 X(a1) = X1 Loop a1 = 0Do While Not EOF(2) Input #2, Y1 a1 = a1 + 1 Y(a1) = Y1 Loop Close #1Close #2PrintPrint Spc(35); 1994年1月至1999年2月社会消费品零售总额（Y:亿元）Print -Print Spc(5); X:;For i = 1 To 11Print Spc(2); X(i);Next iPrintPrint Spc(5); X:;For i = 12 To 22Print Spc(1); X(i);Next iPrintPrint -Print Spc(5); Y:;For i = 1 To 11Print Spc(1); Y(i);Next iPrintPrint Spc(5); Y:;For i = 12 To 22Print Spc(1); Y(i);Next iPrintPrint -End Sub绘制时序图程序：Private Sub Command1_Click()Open C:x.txt For Input As #1Open c:y.txt For Input As #2Dim X(1 To 22), Y(1 To 22), X1!, Y1!, a1%Do While Not EOF(1) Input #1, X1 a1 = a1 + 1 X(a1) = X1 Loop a1 = 0Do While Not EOF(2) Input #2, Y1 a1 = a1 + 1 Y(a1) = Y1 Loop Close #1Close #2Dim i%Form3.Scale (1, 9500)-(30, 3000)DrawWidth = 7Line (3, 3300)-(30, 3300): Line (3, 3000)-(3, 9200)CurrentX = 25: CurrentY = 3800: Print XCurrentX = 4: CurrentY = 9000: Print YFor i = 1 To 21Form3.Line (i + 3), Y(i)-(i + 4), Y(i + 1), vbBlueNext iEnd Sub线性回归结果程序：Private Sub Command1_Click()Open C:x.txt For Input As #1Open c:y.txt For Input As #2Dim X(1 To 22), Y(1 To 22), X1!, Y1!, a1%Do While Not EOF(1) Input #1, X1 a1 = a1 + 1 X(a1) = X1 Loop a1 = 0Do While Not EOF(2) Input #2, Y1 a1 = a1 + 1 Y(a1) = Y1 Loop Close #1Close #2Dim a, b, sumX, sumY, sumXY, sumx2, avrX, avry As DoubleFor i = 1 To 22 求a,b开始sumX = sumX + X(i)sumY = sumY + Y(i)sumx2 = sumx2 + X(i) * X(i)sumXY = sumXY + X(i) * Y(i)Next iavrX = sumX / 22avry = sumY / 22b = (sumXY - 22 * avrX * avry) / (sumx2 - 22 * avrX * avrX)a = avry - b * avrX 求a,b结束Dim ssr, se, e(1 To 22) As Double 求残差平方和For i = 1 To 22 e(i) = Y(i) - (a + b * X(i) ssr = ssr + (Y(i) - (a + b * X(i) 2Next ise = (ssr / 20) 0.5 求回归标准差Dim sst As Double 求离差平方和For i = 1 To 22 sst = sst + (Y(i) - avry) 2Next iDim sse As Double 求回归平方和For i = 1 To 22 sse = sse + (X(i) - avrX) 2Next isse = (b 2) * sseDim r2, sumy2, ar2 As Double 决定系数For i = 1 To 22sumy2 = sumy2 + Y(i) 2Next ir2 = b 2 * (sumx2 - 22 * avrX 2) / (sumy2 - 22 * avry 2)ar2 = 1 - (1 - r2) * 21 / 20 调整决定系数Dim sume1, sume2, p, dw As DoubleFor i = 1 To 21sume1 = sume1 + (e(i + 1) - e(i) 2sume2 = sume2 + e(i) 2Next idw = sume1 / (sume2 + e(22) 2) dw统计量f = r2 / (1 - r2) / 20) f统计量 For i = 1 To 22 样本标准差 sd1 = sd1 + (Y(i) - avry) 2Next isd = (sd1 / 22) 0.5Dim l As Doublel = -11 * (Log(2 * 3.1415926) + 1 + Log(ssr / 22) 对数似然值aic = -2 * l / 22 + 4 / 22 赤池信息准则sc = -2 * l / 22 + 2 * Log(22) / 22 施瓦兹准则Dim ta, tb, stdc, stdx As Doubles2 = ssr / 20 t统计量tb = b / (s2 / (sse / b 2) 0.5ta = a / (s2 * (1 / 22 + avrX 2 / (sse / b 2) 0.5se = (ssr / 20) 0.5stdc = a / tastdx = b / tbPrint Spc(50); 线性回归结果Print -Print -Print Spc(1); variable + Space(7) + coffcient + Space(19) + std.error + Space(19) + t-statistic + Space(10) + Space(4); probPrint Spc(3); c, a, stdc, ta, 0Print Spc(3); x, b, stdx, tb, 0.0000Print -Print -Print Spc(3); R-squared; Spc(19); r2; Spc(7); mean dependent var; Spc(12); avryPrint Spc(3); Adjusted R-squared; Spc(10); ar2; Spc(10); s.d. dependent var; Spc(8); Spc(4); 1366.325Print Spc(3); S.E. of regression; Spc(11); se; Spc(7); Akaike info criterion; Spc(7); aicPrint Spc(3); sum squared resid; Spc(12); ssr; Spc(7); Schwarz criterion; Spc(11); scPrint Spc(3); log likehood; Spc(17); l; Spc(7); F - statistic; ; Spc(15); fPrint Spc(3); Durbin Watson stat; Spc(10); dw; Spc(7); prob(F-statistic); Spc(16); 0.00000Print -Print -PrintPrintPrint Spc(20); 回归方程式：; Y=; a; +; b; *XEnd Sub虚拟变量回归结果程序：Private Sub Command1_Click()Open C:x.txt For Input As #1Open c:y.txt For Input As #2Dim X(1 To 22), Y(1 To 22), X1!, Y1!, a1%Do While Not EOF(1) Input #1, X1 a1 = a1 + 1 X(a1) = X1 Loop a1 = 0Do While Not EOF(2) Input #2, Y1 a1 = a1 + 1 Y(a1) = Y1 Loop Close #1Close #2Dim a, b, sumX, sumY, sumXY, sumx2, avrX, avry As DoubleFor i = 1 To 22 求a,b开始sumX = sumX + X(i)sumY = sumY + Y(i)sumx2 = sumx2 + X(i) * X(i)sumXY = sumXY + X(i) * Y(i)Next iavrX = sumX / 22avry = sumY / 22b = (sumXY - 22 * avrX * avry) / (sumx2 - 22 * avrX * avrX)a = avry - b * avrX 求a,b结束Dim ssr, se, e(1 To 22) As Double 求残差平方和For i = 1 To 22 e(i) = Y(i) - (a + b * X(i) ssr = ssr + (Y(i) - (a + b * X(i) 2Next ise = (ssr / 20) 0.5 求回归标准差Dim sst As Double 求离差平方和For i = 1 To 22 sst = sst + (Y(i) - avry) 2Next iDim sse As Double 求回归平方和For i = 1 To 22 sse = sse + (X(i) - avrX) 2Next isse = (b 2) * sseDim r2, sumy2, ar2 As Double 决定系数For i = 1 To 22sumy2 = sumy2 + Y(i) 2Next ir2 = b 2 * (sumx2 - 22 * avrX 2) / (sumy2 - 22 * avry 2)ar2 = 1 - (1 - r2) * 21 / 20 调整决定系数Dim sume1, sume2, p, dw As DoubleFor i = 1 To 21sume1 = sume1 + (e(i + 1) - e(i) 2sume2 = sume2 + e(i) 2Next idw = sume1 / (sume2 + e(22) 2) dw统计量f = r2 / (1 - r2) / 20) f统计量 For i = 1 To 22 样本标准差 sd1 = sd1 + (Y(i) - avry) 2Next isd = (sd1 / 22) 0.5Dim l As Doublel = -11 * (Log(2 * 3.1415926) + 1 + Log(ssr / 22) 对数似然值aic = -2 * l / 22 + 4 / 22 赤池信息准则sc = -2 * l / 22 + 2 * Log(22) / 22 施瓦兹准则Dim ta, tb, stdc, stdx As Doubles2 = ssr / 20 t统计量tb = b / (s2 / (sse / b 2) 0.5ta = a / (s2 * (1 / 22 + avrX 2 / (sse / b 2) 0.5se = (ssr / 20) 0.5stdc = a / tastdx = b / tbPrint Spc(45); 虚拟变量线性回归结果Print -Print -Print Spc(1); variable + Space(7) + coffcient + Space(13) + std.error + Space(18) + t-statistic + Space(16); prob.Print Spc(3); c, Spc(3); 4565.299, Spc(10); 171.5491, Spc(10); 26.6122, Spc(8); 0.0000Print Spc(3); x, Spc(3); 193.2934, Spc(10); 9.589886, Spc(10); 20.15597, Spc(8); 0.00000Print Spc(3); D1, Spc(3); -824.7766, Spc(10); 172.5291, Spc(10); -4.780506, Spc(8); 0.0002Print Spc(3); D2, Spc(3); -1136.6875, Spc(10); 172.2624, Spc(10); -6.598577, Spc(8); 0.0000Print Spc(3); D3, Spc(3); -1025.1074, Spc(10); 180.1776, Spc(10); -5.689423, Spc(8); 0.0000Print -Print -Print Spc(3); R-squared; Spc(19); 0.964906; Spc(7); mean dependent var; Spc(12); 6020.25Print Spc(3); Adjusted R-squared; Spc(10); 0.956649; Spc(8); s.d. dependent var; Spc(10); 1366.325Print Spc(3); S.E. of regression; Spc(11); 284.482; Spc(7); Akaike info criterion; Spc(9); 0.33593Print Spc(3); sum squared resid; Spc(12); 1375810.23; Spc(7); Schwarz criterion; Spc(11); 14.58390Print Spc(3); log likehood; Spc(17); -152.6953; Spc(7); F - statistic; ; Spc(15); 116.8538Print Spc(3); Durbin Watson stat; Spc(10); 0.423426; Spc(7); prob(F-statistic); Spc(14); 0.00000Print -Print -PrintPrint Spc(20); 回归方程式：; Y=; 4565.299; +; 193.2934; *X; -824.7766*D1-1136.687*D2-1025.107*D3End Sub预测结果分析程序：Private Sub Command1_Click()Open C:x.txt For Input As #1Open c:y.txt For Input As #2Dim X(1 To 22), Y(1 To 22), X1!, Y1!, a1%Do While Not EOF(1) Input #1, X1 a1 = a1 + 1 X(a1) = X1 Loop a1 = 0Do While Not EOF(2) Input #2, Y1 a1 = a1 + 1 Y(a1) = Y1 Loop Close #1Close