问题

I have the represented independently these two plots using R:

#PLOT 1
    x<-250:2500
    #Hsap. Northern European
        a<-dnorm(x,1489,167)
    #Hsap. South African
        b<-dnorm(x,1472,142)

    plot(x,a, type="l", lwd=3, ylim=c(0,1.2*max(a,b,c)), ylab="Probability Density", xlab="Microns")
    lines(x,b, type="l", lwd=3, col="Red")

PLOT 2

    #CUSPAL ENAMEL FORMATION TIME
    x<-0:800
    #Hsap. Northern European
        a<-dnorm(x,447,37)
    #Hsap. South African
        b<-dnorm(x,444,33)

    plot(x,a, type="l", lwd=3, ylim=c(0,1.2*max(a,b,c)), ylab="Probability Density", xlab="Days")
    lines(x,b, type="l", lwd=3, col="Red")
![enter image description here][2]

I would like to merge both using R and obtain an image similar to that shown below. It is interesting to say that I would like as well to stand out the intervals of +- 1SD, to see the overlapping area in the two plots.

Which is the exact code in R to get into success my purpose?

UPDATE

Now, with my data I get the next figure:

As you can see, the overlapping standard deviations are not in the best place. I'd like these overlapping areas were above the normal distributions placed at the X-axis. In this way I could see clearly them.

So the question is, can you write some examples and in that way I learn how I can move these scales to avoid that situation?

In my example I would like to move upwards the right-normal distributions (Y-axis).

x1<-30:200
a1<-dnorm(x1,87,15)
b1<-dnorm(x1,89,13)
c1<-dnorm(x1,92,16)
d1<-dnorm(x1,104,15)

x2<-000:1600
a2<-dnorm(x2,724,66)
b2<-dnorm(x2,724,50)
d2<-dnorm(x2,835,117)

scale<-range(pretty(range(a1,a2,b1,b2,c1,d1,d2)))

remap<-function(x, to, from=range(x)) {
    (x-from[1]) / (from[2]-from[1]) * (to[2]-to[1]) + to[1] 
}

plot(NA, NA, xaxt="n", yaxt="n", type="n", xlim=scale, ylim=scale)
rect(remap(87-15, scale, range(x1)), scale[1],
     remap(87+15, scale, range(x1)), scale[2], col="#ff606025", lty=1)
rect(remap(89-13, scale, range(x1)), scale[1],
     remap(89+13, scale, range(x1)), scale[2], col="#ff606025", lty=2)
rect(remap(92-16, scale, range(x1)), scale[1],
     remap(92+16, scale, range(x1)), scale[2], col="#3dae0025", lty=0)
rect(remap(104-15, scale, range(x1)), scale[1],
     remap(104+15, scale, range(x1)), scale[2], col="#005ccd25", lty=0)

rect(scale[1], remap(724-66, scale, range(x2)),
     scale[2], remap(724+66, scale, range(x2)), col="#ff606025", lty=1)
rect(scale[1], remap(724-50, scale, range(x2)),
     scale[2], remap(724+50, scale, range(x2)), col="#ff606025", lty=2)
rect(scale[1], remap(835-117, scale, range(x2)),
     scale[2], remap(835+117, scale, range(x2)), col="#005ccd25", lty=0)

lines(remap(x1,scale), a1, col="darkred", lwd=3)
lines(remap(x1,scale), b1, col="darkred", lwd=3, lty=3)
lines(remap(x1,scale), c1, col="darkgreen", lwd=3)
lines(remap(x1,scale), d1, col="darkblue", lwd=3)

lines(scale[2]-a2, remap(x2,scale), col="darkred", lwd=3)
lines(scale[2]-b2, remap(x2,scale), col="darkred", lwd=3, lty=3)
lines(scale[2]-d2, remap(x2,scale), col="darkblue", lwd=3)


axis(2); axis(3)
axis(1, at=remap(pretty(x1), scale), pretty(x1))
axis(4, at=remap(pretty(x2), scale), pretty(x2))

Thanks

回答1:

The tricky thing is each plot can only hold one coordinate system. So what we need to do are just scale all the values to the same range, then we can do some fudging on the axis to return to the correct scale. Here's a complete solution

x1<-250:2500
a1<-dnorm(x1,1489,167)
b1<-dnorm(x1,1472,142)

x2<-0:800
a2<-dnorm(x2,447,37)
b2<-dnorm(x2,444,33)

scale<-range(pretty(range(a1,a2,b1,b2)))

remap<-function(x, to, from=range(x)) {
    (x-from[1]) / (from[2]-from[1]) * (to[2]-to[1]) + to[1] 
}


plot(NA, NA, xaxt="n", yaxt="n", type="n", xlim=scale, ylim=scale, xlab="", ylab="")
rect(remap(1489-167, scale, range(x1)), scale[1],
    remap(1489+167, scale, range(x1)), scale[2], col=rgb(0,0,0,.25), lty=0)
rect(remap(1472-142, scale, range(x1)), scale[1],
    remap(1472+142, scale, range(x1)), scale[2], col=rgb(1,0,0,.25), lty=0)

rect(scale[1], remap(447-37, scale, range(x2)),
    scale[2], remap(447+37, scale, range(x2)), col=rgb(0,0,0,.25), lty=0)
rect(scale[1], remap(444-33, scale, range(x2)),
    scale[2], remap(444+33, scale, range(x2)), col=rgb(1,0,0,.25), lty=0)

lines(remap(x1,scale), a1, col="black", lwd=3)
lines(remap(x1,scale), b1, col="red", lwd=3)
lines(scale[2]-a2, remap(x2,scale), col="black", lwd=3)
lines(scale[2]-b2, remap(x2,scale), col="red", lwd=3)
axis(2); axis(3)
axis(1, at=remap(pretty(x1), scale, range(x1)), pretty(x1))
axis(4, at=remap(pretty(x2), scale, range(x2)), pretty(x2))

which produces