问题
I have created a scatterplot3d with a linear model applied.
Unfortunately the results of the LM are subtle and need to be emphasised, my question is how can I extend the LM grid outside of the 'cube'.
Plot:
Code:
Plot1 <-scatterplot3d(
d$MEI,
d$YYYYMM,
d$AOELog10,
pch=20,
grid = FALSE,
color = "black",
xlab="MEI",
ylab="Date",
zlab="AOE Log(10)"
)
fit <- lm(d$AOELog10 ~ d$MEI+d$Rank)
Plot1$plane3d(fit)
Now I guess it might be a variable within lm(), but I cant find anything....
回答1:
To see a larger region, or region of interest, specify the x, y, and z limits in the scatterplot command.
library(scatterplot3d)
d<-data.frame(MEI=runif(200,-3,3),
YYYYMM=runif(200,1,300),
AOELog10=runif(200,1,20),
Rank=runif(200,1,5))
fit <- lm(d$AOELog10 ~ d$MEI+d$Rank)
Plot1 <-scatterplot3d(
d$MEI, d$YYYYMM, d$AOELog10,
pch=20, grid = FALSE, color = "black",
xlab="MEI", ylab="Date", zlab="AOE Log(10)",
main="baseline"
)
Plot1$plane3d(fit)
Plot2 <-scatterplot3d(
x=d$MEI, y=d$YYYYMM, z=d$AOELog10,
pch=20, grid = FALSE, color = "black",
xlab="MEI", ylab="Date", zlab="AOE Log(10)",
xlim = c(-5,5), ylim = c(-50,400), zlim = c(-10,50), # Specify the plot range
main="larger region"
)
Plot2$plane3d(fit)
来源:https://stackoverflow.com/questions/26317538/r-extending-linear-model-beyond-scatterplot3d