0

I am trying to find a way to run a regression on one independent variable with two dependent variables. I have my data organized into a data frame that is 11 observations of 3 variables, the first column containing my independent variable (V1) and the other two containing my dependent variables (V2 & V3).

I have tried the code below.

regression <- lm(binned_data$V2 + binned_data$V3 ~ binned_data$V1)
plot( binned_data$V2 + binned_dataBDI$V3 ~ binned_data$V1, pch =16, cex = 1.0, col = "black", main = "Binned Data and BDI-II Score", xlab = "BDI-II Score", ylab = "Binned Data")
abline(regression)
summary(regression)

I am looking to plot V1 on the x-axis and both dependent variables, V2 & V3, on the y-axis. I also want to include the regression line. I expect there two be 22 data points in total, as there are 11 observations for each dependent variable, but only 11 are plotted.

m4148
  • 91
  • 1
  • 1
  • 3
  • first, your lm should be `lm(as.matrix(data[-1])~iris[,1])` – Onyambu Feb 08 '19 at 00:02
  • Agree that suggestion would succeed but the more general approach would be to use cbind: `regression <- lm( cbind(V2 , V3) ~ V1, data =binned_data)` . See `?lm` – IRTFM Feb 08 '19 at 00:29

1 Answers1

0

Using the iris data as an example, here’s how it can be done using base R. 

mdl <- lm(cbind(Sepal.Length, Sepal.Width) ~ Petal.Length, iris)
plot(Sepal.Length ~ Petal.Length, iris, ylim = c(0, 9))
points(Sepal.Width ~ Petal.Length, iris, pch = 3)
abline(mdl$coefficients[, 1])
abline(mdl$coefficients[, 2])
summary(mdl)

example plot

Nick Kennedy
  • 12,510
  • 2
  • 30
  • 52