0

I have an assignment talking about plotting the data points in R3.

The data are shown below:

no x1 x2 x3 y

1 9 12 7 101

2 3 6 6 66

3 13 14 1 85

4 10 12 7 103

5 4 4 3 47

6 -5 -1 4 19

7 3 8 6 72

8 11 1 6 67

I tried to plot them using the scatter plot, three set of dots (i.e. x1 vs y, x2 vs y and x3 vs y) were on the same graph but the tutor said that it was incorrect.

Does anyone tell me how to plot these data?

Thank you very much!

chrisych
  • 3

1 Answers1

0

Looking at your data $$\left( \begin{array}{cccc} x_1 & x_2 & x_3 & y \\ 9 & 12 & 7 & 101 \\ 3 & 6 & 6 & 66 \\ 13 & 14 & 1 & 85 \\ 10 & 12 & 7 & 103 \\ 4 & 4 & 3 & 47 \\ -5 & -1 & 4 & 19 \\ 3 & 8 & 6 & 72 \\ 11 & 1 & 6 & 67 \end{array} \right)$$ and using a multilinear regression, the data correspond exactly to $$y=12+2x_1+3x_2+5x_3$$ which is the equation of a straight line in a four dimension space.

Now, how to plot them on a piece of paper ?

Claude Leibovici
  • 260,315
  • How did you obtain the equation y = 12 + 2x1 + 3x2 + 5x3? The question (about the data mining and the principal component analysis) is asked to plot the data points in R3 (i.e. 3D) but not 4D. Should I use R or Weka to plot these data? – chrisych Oct 09 '16 at 08:26
  • @chrisych. I suspected that the model would be linear. So, I just applied an standard multilinear regression from which I got the coefficients. Now, I do not know the tools you speak about (since I am almost blind, I do not care too much about them). If you have appropriate tools, may be an idea could be to make a 3D contour plot, each level corresponding to a value of $y$. – Claude Leibovici Oct 09 '16 at 08:35
  • @chrisych. You are very welcome ! – Claude Leibovici Oct 09 '16 at 08:38