CODEWITHSUNDEEP
rplotregression
Bobe Kryant
Bobe Kryant

Reputation: 2100

Plot multiple polynomial regression curve

I am trying to plot only a few regression lines and not any of the points. (No fitted, because I have over 7 thousand points.) I know how to do this with linear regressions, but not with polynomial regression. My data is here. With a few linear regressions:

plot_data=read.csv("plot_data.csv") #read data

#linear regressions
Off_linear=lm(Z_Salary~OBPM,data=plot_data) 
Def_linear=lm(Z_Salary~DBPM,data=plot_data)
Tot_linear=lm(Z_Salary~BPM,data=plot_data)   

#try to plot. This works. Not sure how to add legend
termplot(Def_linear, ylab='Z_Salary',xlab='BPM',ylim=c(-2, 2))
abline(Off_linear)
abline(Tot_linear,col='blue')

Plot

However I cannot do this if I try to create polynomial regressions. I want the same kind of plot with these regressions, but termplot does work with independent variables together and does them separately.

Off_exp=lm(Z_Salary~OBPM+I(OBPM^2),data=plot_data)
Def_exp=lm(Z_Salary~DBPM+I(DBPM^2),data=plot_data)
Tot_exp=lm(Z_Salary~BPM+I(BPM^2),data=plot_data)

Upvotes: 0

Views: 1274

Answers (1)

lrnzcig
lrnzcig

Reputation: 3947

@Dason already gave you the hint. Find below some code to make it work.

Off_exp=lm(Z_Salary~OBPM+I(OBPM^2),data=plot_data)
x=seq(from=range(plot_data$OBPM)[1], to=range(plot_data$OBPM)[2])
y=predict(Off_exp, newdata=list(OBPM=x))
plot(x, y, type="l", col="blue", xlab="OBPM", ylab="Z_Salary")
legend("topleft", legend="d2", col="blue", lty=1, title="polys")

It would look something like this: resulting plot

Hope it helps.

Upvotes: 1

Related Questions