今年年初起开始学习天体物理学(虚构的,其实就是想测试我新上传的红色好看的头像怎么不显示)。
霍金的《十一维空间》;
记得以前和几个物理学家……,编不下去了。哈哈。
> options(width=80)> library(BsMD)> data(BM86.data,package="BsMD")> print(BM86.data)
X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11 X12 X13 X14 X15 y1 y2 y3 y41 -1 -1 1 -1 1 1 -1 -1 1 1 -1 1 -1 -1 1 0.23 43.7 14.0 0.082 1 -1 -1 -1 -1 1 1 -1 -1 1 1 1 1 -1 -1 0.30 40.2 16.8 0.043 -1 1 -1 -1 1 -1 1 -1 1 -1 1 1 -1 1 -1 0.52 42.4 15.0 0.534 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 1 1 1 1 0.54 44.7 15.4 0.435 -1 -1 1 1 -1 -1 1 -1 1 1 -1 -1 1 1 -1 0.70 42.4 27.6 0.316 1 -1 -1 1 1 -1 -1 -1 -1 1 1 -1 -1 1 1 0.76 45.9 24.0 0.097 -1 1 -1 1 -1 1 -1 -1 1 -1 1 -1 1 -1 1 1.00 42.2 27.4 0.128 1 1 1 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 0.96 40.6 22.6 0.369 -1 -1 1 -1 1 1 -1 1 -1 -1 1 -1 1 1 -1 0.32 42.4 22.3 0.7910 1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 1 1 0.39 45.5 17.1 0.6811 -1 1 -1 -1 1 -1 1 1 -1 1 -1 -1 1 -1 1 0.61 43.6 21.5 0.7312 1 1 1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 0.66 40.6 17.5 0.0813 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 0.89 44.0 15.9 0.7714 1 -1 -1 1 1 -1 -1 1 1 -1 -1 1 1 -1 -1 0.97 40.2 21.9 0.3815 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1 -1 1.07 42.5 16.7 0.4916 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1.21 46.5 20.3 0.23Saturated linear models for each of the responses are fitted and the estimated coefficients arepresented in the table below. The lm calls, not displayed here, produce the advance.lm, . . . ,yield.lm objects used in the next subsections.advance shrinkage strength yield(Intercept) 0.70 42.96 19.75 0.38X1 0.03 0.06 -0.30 -0.10X2 0.13 -0.07 -0.20 -0.01X3 -0.01 0.15 -0.30 0.00X4 0.25 0.07 2.30 -0.04X5 0.00 0.20 0.45 0.02X6 -0.01 -0.01 -0.10 -0.03X7 0.00 0.19 -0.15 0.07X8 0.07 0.20 -0.60 0.14X9 0.01 -0.03 0.35 -0.08X10 0.00 0.21 0.05 -0.13X11 0.01 0.06 0.15 -0.05X12 0.02 0.06 -2.75 -0.01X13 0.01 -0.19 1.90 0.00X14 -0.01 1.07 0.05 0.06X15 0.01 1.55 -0.30 0.01For each of the experiments the 16 runs are used on the estimation of the 15 contrasts and theconstant term. Thus the need of graphical aims to determine which are likely active contrasts.2.1 Daniel PlotsDaniel plots, known as normal plot of effects, arrange the estimated factor effects in a normalprobability plot; those factors “out of the straight line” are identified as potentially active factors.See for example, Daniel (1976) for different applications and interpretations.
DanielPlot produces normal plot of effects. The main argument of the function is an lm object,say, lm.obj. The function removes the constant term (Intercept) if it is in the model. Factoreffects, assumed as 2*coef(lm.obj) are displayed using the qqnorm function. See the help pagesfor details.2.1.1 Box et al. 1986: Example 1By default DanielPlot labels all the effects, as show in figure a). This example shows how to labelonly some particular factors for clarity, as exhibited in figure b). The corresponding linear modeladvance.lm was already fitted at the beginning of the section.> par(mfrow=c(1,2),mar=c(3,3,1,1),mgp=c(1.5,.5,0),oma=c(0,0,0,0),+ xpd=TRUE,pty="s",cex.axis=0.7,cex.lab=0.8,cex.main=0.9)> DanielPlot(advance.lm,cex.pch=0.8,main="a) Default Daniel Plot")> DanielPlot(advance.lm,cex.pch=0.8,main="b) Labelled Plot",pch=20,+ faclab=list(idx=c(2,4,8),lab=c(" 2"," 4"," 8"))
BsMD.pdf (rstudio.com)
上传头像。




雷达卡






京公网安备 11010802022788号







